PB85-17':694
REVIEW OF DERMAL ABSORPTION
U.S Environmental Protection Agency
Washington,  DC
Oct 84
                     U.S. DEPARTMENT OF COMMERCE
                   National Technical Information Service
                                  NITS

-------

-------
                                   DISCLAIMER
     This report has been reviewed in accordance with U.S. Environmental
Protection Agency policy and approved for publication.  Mention of trade  names
or commercial products does not constitute endorsement or recommendation  or
use.
                                       ii

-------
                                    CONTENTS

Foreword	vii
Abstract	viil
1.   Introduction	1
2.   Conclusions	2
3.   Recommendations	4
4.   Skin as a Barrier to Absorption	7
     4.1  Structure of the Skin
     4.2  Factors Affecting Absorption
5.   Techniques for Measuring Absorption	15
     5.1  In Vivo Human Studies
     5.2  In Vivo Animal Studies
     5.2  Tn Yftro
6.   Theoretical Treatment of Dermal Absorption	38
     6.1  Pick's Law Applied to Dermal Absorption
     6.2  Non-Steady State Diffusion
     6.3  Calculation of kp and Km
     6.4  Prediction of Permeability for Some Alcohols and Steroids
7.   Dermal Absorption in Exposure Assessments	58
     7.1  Current Agency Practice
     7.2  Examples of Dermal Intake Calculations
          7.2.1  Dermal Exposure from Swimming
          7.2.2  Pesticide Application
          7.2.3  Treated Field Reentry
          7.2.4  Hypothetical Dermal Intake of PCBs
8.   Analysis/Future Research Needs	69
9.   References	74
                                      iii

-------
                                 LIST OF TABLES

                                                                        Page

1.   Relationship Between Dose and Efficiency of Absorption ....... 10

2    a. Ratio Comparing Absorption of Various Sites to Forearm ..... 12
     b. Penetration Indices for Five Anatomic Sites ........... 12
     c. Body Surface Areas for Five Anatomic Sites ........... 12

3.   Percutaneous Absorption of Steroids in Man ............. 16

4.   Percutaneous Absorption of Some Organic Compounds in Man ...... 19

5.   Percutaneous Absorption of Some Pesticides and Herbicides in Man. . 22

6.   In Vivo Percutaneous Absorption by Rat, Rabbit, Pig, and Man. ... 24

7.   In Vivo Percutaneous Absorption of Several Pesticides by Rabbit,
     TTg, Squirrel Monkey, and Man ................... 26
8.   Geometric Means of Percentage   C Recovered in Various Fractions
     at 5, 15, and 60 Min Post Application ................ 27

9.   Comparison Between Physical  Parameters and Geometric Means of
     Rate of Penetrations ........................ 28

10.  Comparison of Human In Vivo and In Vitro Absorption ......... 32

11.  Comparison of Human In Vivo and In Vitro Absorption Using
     Refined Techniques ......................... 34

12   a. Percutaneous Absorption of Acetyl salicylic Acid in Rats ..... 36
     b. In Vivo vs. In Vitro Percutaneous Values with Those of Other
        "STudPTe? ............................. 36
     c. Comparison of Permeation Values with Those of Other Studies . .  .36

13.  Membrane Permeability and Partition Coefficients ..... ..... 45

14.  Permeability Constants and Partition Coefficients .......... 47

15.  Steroid Fluxes ................. . ......... 49

16.  In Vitro Permeability of Aliphatic Alcohols Through Human
     "Epidermis .............................. 54

17.  a. In Vitro Permeability of Steroids Through Human Epidermis ...  .55
     b. Trii Vitro Permeability of Steroids Through Human Epidermis ...  .56
                                       IV

-------
                          LIST OF TABLES (Continued)

                                                                        Page


18.  a. Relative Contribution to Total  Dermal  Exposure of Body Areas
        to Pesticides as Studied by the California Department  of Food
        and Agriculture	63

     b. Relative Contributions to Total Dermal  Exposure of Body Areas
        by Job Activity	63

19.  Worker Reentry Exposure 	 	 65

-------
                                LIST OF FIGURES


                                                                     Page

1.  Skin Cross Section	7

2.  Schematic Representation of a Diffusion Cell  with Top Open to
    the Ambient Environment.	30

3.  Diagram Showing the Concentration Profile Across Stratum Corneum
    During Steady-State Absorption	39

4.  The Pattern of Changing Absorption Rate for Small Amounts of
    Penetrant Per Unit Area of Skin	42

5.  The Extrapolation to Zero Absorption of the Steady-State,
    Linear Region of the Graph of Total Amount Absorbed Versus
    Time Yields the Lag Time	48

6.  Dependence of the Permeability Constant on the Stratum Corneum/
    Water Partition Coefficient	53
                                         VI

-------
                                    FOREWORD


     The Exposure Assessment Group of EPA's Office of Research and Development

has three main functions:  1) to conduct exposure assessments; 2)  to review

assessments and related documents; and 3) to cievelop guidelines for Agency

exposure assessments.  These exposure assessments are critical to  the evaluation

of the public health risks that are presented by toxic substances.  The activities

under each of these three functions are supported by and respond to the needs of

the various EPA program offices.  In relation to the third function, the Exposure

Assessment Group conducts projects for the purpose of developing or refining

techniques used in exposure assessments.

     The current project was initiated to provide a review of human dermal

absorption since this is a mechanism by which toxicants can enter  the body.

To properly conduct an exposure assessment, all  routes of exposure should be

determined.  Besides reviewing current dermal absorption studies,  future

research needs are addressed which could further elucidate the methods needed

for dermal exposure calculations.  The Exposure  Assessment Group eventually

hopes to provide similar reviews for all routes  of exposure.

                                                   James W. Falco, Ph.D., P.E.
                                                   Director
                                                   Exposure Assessment Group
                                      vn

-------
                                 ABSTRACT

     Dermal absorption is one of the three main routes of human exposure.   This
report is a review of the current dermal  absorption  literature with an  emphasis
on applications for exposure assessments.  The structure of the skin is described
in the first technical  chapter as well  as factors  such as occlusion or  abrasion
which affect absorption rates.  The next  chapter presents comprehensive tables
of in vivo human dermal  studies, in vivo  animal  studies, and in vitro results
using either excised animal  or human skin.  Theoretical treatments of dermal
absorption are discussed next, including  kinetic expressions for various dermal
exposure situations and ways to predict absorption using partition coefficients.
Following this theoretical  review some  calculations  of typical  dermal exposure
scenarios are presented such as chemicals in swimming pools or reentry  into
pesticide treated fields.  Lastly, an analysis of  some data gaps in dermal
absorption knowledge is coupled with suggestions for future research.
                                      viii

-------
                                1.  INTRODUCTION

     Three main routes of human exposure to toxic  compounds  have  been  identified,
namely inhalation, ingestion, and dermal penetration.   It. is recognized that
for certain scenarios, such as farm worker reentry following pesticide application
or occupational handling of pure liquids or solids, exposure via  the dermal
route is the most critical  exposure pathway.  In the past, when the Agency was
faced with estimating the dermal absorption of a compound without the support
of scientific studies, a value of 10 percent absorption was  assumed.  Recently,
the Office of Pesticide Program's Scientific Advisory  Panel  stated that if no
literature was available on a compound to show otherwise, then a  value of 100
percent of the penetrant on the" skin should be used to represent  the dermal
absorption.  This value of 100 percent is a conservative approach and  leads to
the overestimation of actual dermal absorption for most compounds.  The intent
of this document is to review the current state of knowledge of dermal  exposure
focusing on ways to assign more realistic values for absorption  (i.e., <100
percent).

-------
                                2.  CONCLUSIONS

     The review of the literature on dermal  absorption presented  in  this  report
has brought out a number of factors which help to explain how dermal  absorption
of a penetrant occurs.  It has been shown that the thin outermost layer of
skin, the stratum corneum, is the rate limiting membrane for diffusion.   For
most penetrants, absorption through the general  skin  surface is the  preferred
route over "holes" in the skin caused by hair follicles and sweat ducts.   The
site of dermal exposure is important too as  studies have shown up to  a ten-fold
difference in absorption rate depending on where on the body a penetrant  is
applied.  There are a number of other factors which can affect the amount  of
penetrant absorption such as condition of the skin, occlusion or  covering  of
the applied dose, frequency of application,  metabolism of a penetrant by the
skin, and the solvent or formulation used to deliver a penetrant  to  the skin.
     In vivo studies in man have been done for approximately 50 chemicals,
mostly pesticides and steroids.  Studies have shown up to a 5-fold variation
in skin absorption between subjects.  Toxicity concerns prohibit  further testing
in man, thus in vivo studies are done on animals.  There is a large  variation
in penetrant absorption in the animal species tested  with the rhesus  monkey
and miniature swine appearing to be closest  to man.  In vitro studies using
either excised human or animals skin give qualitative agreement with  in vivo
results and show the potential for quantitative use.
     Attempts have been ^ade tc predict dermal  absorption from partition
coefficients, particularly with octanol/water or olive oil/water.  It has  been
found, however, that a compound has to be both lipid  soluble and  water soluble
to be a good penetrant.  The solubility of the penetrant within the  stratum

-------
corneum relative to its solubility in the vehicle (i.e.,  the  membrane  partition
coefficient, Km) is a better indicator of permeability  than octanol/water or
olive oil/water partition coefficients.  Values for K^  have been  used  to
predict the permeability of the linear primary alcohol  series with  good success.
     The absorption rate is just one factor needed to calculate dermal  intake
for an exposure assessment.  In addition, one  needs to  know the duration of
the exposure per each event, the frequency of  events expressed in number of
exposures per some unit time, the ambient concentration of the penetrant as a
function of time plus the medium or vehicle in which it is applied  to  the skin,
and the area of exposed skin.  The uncertainty contained  in these factors is
large and may necessitate the use of approximations.

-------
                              3.   RECOMMENDATIONS


     One stated goal  of this review is  to work  towards  assigning  a  more  realistic

value for the percent absorption  of an  untested penetrant  rather  than  the

current conservative  approach of  using  100 percent.   Since  so  few compounds

have been tested in vivo in man and toxicity concerns prevent  the testing  of

many compounds, it is necessary to study  penetrants  by  in  vivo animal  tests, _i£

vitro animal or human tissue tests, or  by calculating potential absorption from

partition coefficients or other physical  parameters.   In order to work towards

the goal of being able to assign  a value  of less than 100  percent for  the

absorption rate of an untested penetrant, it is recommended that:

     -  A data base be established of Km  and diffusivity values for a  number of
        chemicals representing the various chemical  classes.

     -  Quantitate the factors affecting  absorption  such as vehicle, site  of
        application,  condition of the skin, occlusion,  frequency  of application,
        and metabolism.  In particular, the effect of the  carrier solvent
        (vehicle) on  the rate of  absorption should be determined  by measuring
        the rate for a pure penetrant and then  dissolving  the  penetrant  in a
        number of solvents (such  as those commonly found in industry).

     -  Develop kinetic expressions to  represent the various dermal  exposure
        situations that occur such as finite amount  of  a fast  penetrant, finite
        amount of a slow penetrant, excess of a penetrant  where the diffusion
        through the stratum corneum is  not rate limiting,  buildup of penetrant
        in blood such that the concentration (C0) is fQ, etc.   Also, it  is
        necessary to  develop procedures for identifying which  kinetic  expression
        is appropriate for use with a given penetrant.

     -  Determine the most suitable animals species  (probably  the rhesus monkey
        or miniature swine, or rat because of ease of handling and  economy) and
        do all animal in vivo work only on th.is species.   The  relationship
        between in vivo in man versus in  vivo in animals should be  established
        by comprehensive testing  of penetrants  that  have already  been  studied
        in man.

     -  For in vitro work, decide from  what part of  the body tissue should be
        taken (forearm, back, stomach).  Investigate problem of low water

-------
        solubility penetrants by using solvent/water or straight  solvent  for
        the receptor cell  (the bottom part  of  the  diffusion cell  apparatus).
        Work on making this method more quantitative so that  in vitro  results
        can substitute for in vivo results. Also, the  relative permeability of
        all body tissue that could potentially be  exposed should  be measured
        and correlated to  the body tissue area selected for the in vitro  studies.

     -  Establish the relationship between dermal  absorption  and  the various
        partition coefficients, particularly Km and Koctanol»  and see  if  tne
        methods used to calculate partition coefficients  such  as  structure-
        activity, number and kind of functional  groups, etc.,  can be used to
        calculate dermal  absorption directly.

     -  Determine the feasibility of grouping  penetrants into  a numerical  system
        like 100-10-1-.!% absorption depending on  in vivo or  in vitro  studies,
        or physical parameters like Km or other partition coeTficients.

     It may be possible to develop a model  which includes all  possible parameters

affecting dermal  intake based on the results of the recommended studies.   One

could store in the model  such factors as:  skin area for total body and  for

each body region depending on age, sex, and body weight;  thickness of  skin for

different body regions; differences in percent or  rate  of absorption (based on

some unity scale) for various body sites; kinetic  expressions  to  determine the

absorption for a range of scenarios; representative K^  and diffusion constants

for chemical  classes and/or functional  groups, etc.  Then one  could enter data

specific for a particular exposure, such as penetrant partition coefficients,

volatility, concentration, and number and kind of  functional  groups, vehicle

(including any binding to soil, or other substrates), contact  time, skin  area

plus where on the body, etc., and have the model calculate the dermal  intake.

-------
                      4.  SKIN AS A BARRIER  TO  ABSORPTION

4.1  STRUCTURE OF SKIN
     Skin is one of the largest organs  of the body  comprising  approximately 10
percent of the normal  body weight.  The skin consists  of two different  layers
with the thinner outer layer known as epidermis and the thicker,  inner  layer
as dermis (Fig. 1).  Although skin thickness varies with location in  humans,
the epidermis is approximately 0.1 mm and the dermis 2-4 mm thick (Rongone,
1983).  The outermost layer of the epidermis is called the stratum corneum and
is from 10 to 50 urn thick.  In this document we are mainly concerned  about the
stratum corneum as it has been shown that the stratum  corneum  is  at least
three, and frequently as much as five,  orders of magnitude less permeable to
most substances as the dermis (Michaels et al., 1975).  Also,  the permeability
of the entire epidermis is indistinguishable from that of the  stratum corneum
alone.  Thus, the skin can be thought of as  a three layer laminate of stratum
corneum, remainder of the epidermis, and dermis with permeation occurring by
diffusion through the three layers in series.
     The stratum corneum is a heterogeneous  structure  containing  about  40%
water, about 40% protein (primarily keratin), and about 15 to  20% lipids  (mainly
triglycerides, free fatty acids, cholesterol, and phospholipids)  (Michaels et
al., 1975 and Anderson and Cassidy, 1973).  The stratum corneum can absorb up
to six times its own weight in water, and in its fully hydrated state,  its
permeability to water and other low molecular weight penetrants is increased
(Scheuplein, 1967).  The lipid components of the stratum corneum  may  be the
chief reason for its uniquely low permeability  as when this tissue is extracted

-------
                                                                     Epidermis
 Living
dividing
 cells
  Blood
  vessel
                                  Subcutaneous
                                    .tissue
                      Figure 1.   Skin Cross Section

                      (By Permission  from the  New York Times)

-------
with a fat solvent and then rehydrated, the water permeabil ity of the stratum
corneum and its permeability to larger molecules is greatly  enhanced (Scheuplein,
1967).
     A recent study by Eli as, et al. (1981) related dermal  absorption to stratum
corneum structure and lipid composition.  The extreme sensitivity of the
permeability barrior to damage by lipid solvents suggest that lipids are
important determinants of skin penetration.  In  addition to  lipids,  several
other stratum corneum structural parameters, including thickness, number of
cell layers, and geometric organization could determine stratum corneum
permeability.  The authors correlated the in vitro penetration of water  and
salicylic acid across leg and abdominal  stratum  corneum with both the lipid
composition and the structure of the same samples.  One finding from this study
is that there is an apparent noncorrelation of penetration of both substances
with either stratum corneum thickness or number  of cell layers.  The results
suggest that dermal absorption of both substances  correlate  with the lipid
content by weight of the sample.  Also, the study  suggests  that relatively
small  inherent variations in lipid concentration may explain observed differences
in permeation across different topographic regions.  The authors predict that
regions of high permeability such as palms and soles would have a low lipid
weight percent, whereas those of low permeability  such as facial or  perineal
stratum corneum, would have a relatively high lipid weight percent.
     The role of skin appendages such as hair follicles and  sweat ducts  on  skin
permeability has been studied extensively.  One  would think  that these "holes"
in the skin would facilitate the passage of a penetrant; however, their  total
area is relatively small  and, for most penetrants, absorption through the
general skin surface is the preferred route (Dugard, 1983).   In the  case of  some

                                      -8-

-------
molecules that penetrate the bulk of the stratum corneum  slowly,  such  as
electrolytes and polar molecules  with three  or  more  labile  polar  groups, the
route through follicles and ducts may predominate (Scheuplein,  1980).

4.2  FACTORS AFFECTING ABSORPTION
     A general definition of percutaneous absorption can  be given as the
penetration of substances from the outside into the  skin  and through the skin
into the blood stream.  Schaefer and Schalla (1980)  have  broken this process into
individual  parts as:  (1) penetration is considered  to  be the process  of entrance
into 1 layer; (2) permeation is the migration through 1 or  several  skin layers;
(3) resorption is the uptake by the cutaneous microcirculation; and  (4) absorption
is the sum of all these processes.
     There are a number of parameters which  can affect  the  amount of penetrant
absorption.  The concentration of applied dose  and surface  area are the two
most important factors, with the greatest potential  for absorption  occurring
when a high concentration of a penetrant is  spread over a large surface area of
the body (Wester and Noonan, 1980a).  The relationship  between  the  concentration
of applied dose and the efficiency of absorption is  not necessarily a  linear
one as studies have shown the efficiency to  decrease with increasing dose  (Table
1).  The site of application was  the forearm for both man and rhesus monkey in
Table 1; note the similarity in percent absorbed. The  absolute dose absorbed
increases, of course, as the dose applied is increased.
     The solvent or formulation used to deliver a penetrant to  the  skin (vehicle)
can have a decided impact on the  efficiency  of  absorption.   The ability of a
compound to penetrate the skin and exert its effect  is  dependent  on two
consecutive physical events.  The compound must first Diffuse out of the vehicle
                                      -9-

-------
        TABLE 1.  RELATIONSHIP BETWEEN DOSE AND EFFICIENCY OF ABSORPTION
Penetrant
Hydrocortlsone
Benzole Add
Testosterone
Dose (ug/cm2)
4
40
4
40
2000
4
400
Percent
Rhesus
2.9
2.1
59.2
33.6
17.4
18.4
2.2
of Dose Absorbed
Man
1.9
0.6
42.6
25.7
14.4
13.2
2.8
from Wester and Noonan 1980a
                                       10

-------
to the skin surface, and then must penetrate the skin  in  route to  the  site  of
action (Ostreng et al., 1971).  If the membrane diffusion constant for the
penetrant and the thickness and solvent properties  of  the membrane are unchanged
by the nature of the external vehicle, then the rate of absorption is  proportional
to the chemical potential in the vehicle.   When the chemical  potential  and
penetrant concentration are linearly related, Pick's law of diffusion  (see
Chap. 6) is obeyed for the vehicle (Dugard, 1983).   Some solvents, such as
DMSO, actually dissolve lipids, destroying the barrier function  and carrying
the penetrant along with it into the body.  The effect of a solvent on the  rate
of absorption of a penetrant should be studied further.
     The site of application of a penetrant can also be an important fact in
the efficiency of absorption.  Maibach and Feldmann (1971) have  measured the
absorption of several compounds at various sites on the body.  Their results
show a wide range of values with the palm  allowing  approximately the same
penetration as the forearm, the abdomen and dorsum  of  the hand having  twice the
penetration of the forearm, the scalp, angle of the jaw,  postauriacular area,
and forehead having four-fold greater penetration,  and the scrotum allowing
almost total absorption (Table 2a).
     Guy and Maibach (1984) divided the body into five regions for the purpose
of calculating correlation factors for use with forearm penetration data:
genitals, arms, legs, trunk, and head.  Table 2b gives "penetration indices" or
the ratio of skin penetration for one of the five anatomic sites divided by
skin penetration for the forearm.  These penetration indices  are derived from
hydrocortisone skin penetration data and from absorption  results using the
pesticides malathion and parathion.  The authors show  the relative proportion
of body area and actual skin area for the  five anatomic sites  in Table 2c.
                                      -11-

-------
                TABLE 2a.  RATIO COMPARING ABSORPTION OF VARIOUS SITES  TO FOREARM
        Site
             Parathion
        From Malbach and Feldmann, 1971.
Mai athi on
Hydrocortlsone
Forearm
Palm
Foot, ball
Abdomen
Hand dorsum
Scalp
Jaw angle
Forehead
Scrotum
1.0
1.3
1.6
2.1
2.4
3.7
3.9
4.2
11.8
1.0
0.9
1.0
1.4
1.8
...
...
3.4
...
1.0
0.8
—
...
...
3.5
13.0
6.0
42.0
                     TABLE 2b.  PENETRATION INDICES FOR FIVE ANATOMIC  SITES
        Site
            Penetration Index based
              on hydrocortlsone
         Penetration Index based
           on pesticide data
Genitals
Arms
Legs
Trunk
Head
40
1
0.5
2.5
5
12
1
1
3
4
        From Guy and Malbach, 1984 (see original  article for the number of footnotes
        describing how these Indices were calculated)
                     TABLE 2c.  BODY SURFACE AREAS FOR FIVE ANATOMIC SITES
                       Adult8                      Ch1ldb                   Neonatec

               Body area (X) Area (cm2)   Body area (S)   Area (cm2)    Body area  (*)    Area  (cm2)
Genitals
Arms
Legs
Trunk
Head
1
18
36
36
9
190
3420
6840
6840
1710
1
19
34
33
13
75
1425
2550
2475
975
1
19
30
31
19
19
365
576
595
365
Total
           19000
    7500
             1920
Fran Guy and Malbach, 1984
aAdult:  weight =
bChild:  weight =
cNeonate:  we1ght
70 kg; height - 1.83m
19 kg; height = 1.10m
= 3 kg; height - 0.49m
                                               -12-

-------
     The condition of the skin, such  as  loss  of barrier  function  of the stratum
corneum through disease or damage,  will  also  affect  absorption.   Absorption  can
be virtually 100% if all  barrier function  is  removed.  Occulusion or covering
of the applied dose as with bandaging or putting on  clothing  after a dermal
application will  increase absorption. Occlusion changes the  hydration and
temperature of the skin and also prevents  the accidental wiping off or evapora-
tion of an applied dose (Wester and Noonan,  1980a).
     The frequency of dermal applications  is  also a  factor  which  could affect
absorption.  In one study the absorption of a single application  of a high
concentration of penetrant was greater than when the equivalent concentration
was applied in equally divided doses  (Wester  et al., 1977).
     In another study the effect of repeated  applications of  a penetrant was
tested.  The absorption on the 8th  day of  application  of the  same penetrant
dose was found to be significantly  higher  than on the  first day.   The authors
suggest that the initial  applications of penetrant altered  the barrier function
of the stratum corneum, resulting in  increased absorption for subsequent
applications (Wester et al., 1980b).
     The skin contains many of the  same  enzymes as the liver, thus penetrant
metabolism by the skin could affect absorption.  The metabolizing potential  of
skin has been estimated to be about 2% of  the liver  with most of  the enzyme
activity localized in the epidermal layer  (Pannatier et  al.,  1978).  The slower
a penetrant is absorbed through the skin the  greater the opportunity for some
metabolism to occur.  In vitro penetrant absorption  studies using excised human
skin will not show the results of any potential metabolism  and, thus, may not
reflect the actual in vivo absorption for  some penetrants.  The metabolism
potential of skin should be studied further.
                                      -13-

-------
                CHAPTER 5 - TECHNIQUES FOR MEASURING ABSORPTION
5.1  IN VIVO
     Percutaneous absorption In vivo is usually determined by measuring  radio-
activity in excreta following the topical  application of  a labeled  compound.
The penetrant under study is labeled with carbon 14 or tritium and  following
application, the total  amount of radioactivity excreted in urine  or urine  plus
feces is determined.  The radioactivity in the excreta can be a mixture  of the
parent compound and any metabolites.  The amount of radioactivity retained in
the body or excreted through expiration or sweat can be determined  by  measuring
the amount of radioactivity excreted following an intravenous (i.v.) injection;
the fraction determined by i.v. dose is then used to correct the  amount  of
radioactivity found after topical  administration (Wester  and Maibach,  1983).
     Another in vivo method of measuring absorption is the detection of  plasma
radioactivity following topical administration.  This method may  be difficult
to apply because penetrant concentrations in blood after  topical  administration
are often very low.  Determining the loss of material from the surface as  it
penetrates the skin is another way to measure in vivo absorption.  It  is assumed
that the difference between applied dose and residual dose is the amount of
penetrant absorbed.  The difficulties inherent in skin recovery,  volatility of
penetrant, and errors associated with using the difference between  amount
applied and amount left make this method less quantitative.  Biological  or
pharmacological  responses, such as vasodialation, have also been  used  to estimate
absorption for a limited number of compounds (Stoughton et al., 1960)
     The main source of human in vivo data is the work of Feldmann  and Maibach
                                      -14-

-------
(1969 and 1970).  These authors studied the percutaneous  absorption  of  some
steroids, pesticides, and organic  compounds using  radiolabeled  (Cl4) doses
administered intravenously and topically to human  volunteers.   The intravenous
dose was used to correct for the amount of radiolabeled penetrant that  is
percutaneously absorbed but not recovered in urine samples  (i.e., excreted in
feces or retained in the body).  The topical dose  of 4 ug/cm2 was dissolved
in acetone and applied to a 13 cm^ circular area of the ventral  forearm.
All urine was collected for five days divided into suitable time periods so
that either the absorption rate (%/hr) or total absorption  (% of dose)  could
be calculated.  The skin sites were not protected  and not washed for 24 hours.
Tables 3-5 are a compilation of the 49 penetrants  studied.
     Table 3 shows the percutaneous penetration of some steroids in  man.  For
the intravenous dose, one microcurie (14C) of the  steroid was dissolved in 5 to
20 ml  of saline (or saline plus ethanol) and injected. The half life,  which is
an indicator of the relative speed of elimination  of the  injected dose, was
obtained from plotting the amount  of l^C excreted  divided by the time period
expressed as the percent of dose administered versus the  collection  t.ime.  The
values for percent recovery are given in the first part of  Table 3 with
testosterone showing almost complete elimination from the body  and fluocinolone
yielding only one-third of the injected dose.  Cortisone, corticosterone, 17-OH
desoxycorticosterone, desoxycorticosterone, and 17-OH progesterone are  assumed
to be excreted similarly as hydrocortisone.  Likewise, the  value for testosterone
was used for the acetate and propionate, plus dehydrocortisone  and androstenedione,
The second part of Table 3 shows absorption after  topical administration.  The
                                      -15-

-------
                      TABLE 3.   PERCUTANEOUS ABSORPTION OF  STEROIDS  IN MAN

I.V. Administration —
           Steroid.-                    8 Recovered                    Half Life (hrs)
Hydrocortlsone
Hydrocortlsone acetate
Estradlol
Testosterone
Progesterone
Fluoclnolone acetonlde
Dexamethasone
65.4
68.9
51.6
92.1
68.7
37.0
47.4
5
5
8
4
4
5.5
4
Topical Administration*.-

                                                                        Total
   Steroid	Absorption  Rate  (t/hr)	    Absorption       No.
                 Time (hrs)                                         7
                   0-12     12-24	24-48   48-72   72-96   96-120   % of dose  S.O.
Subjects
Hydrocortlsone     .005     .023     .019     .018    .016     .010      1.87     1.59   15

Hydrocortlsone
 acetate           .020     .069     .032     .024    .015     .008      2.65     1.80    6

Cortisone          .015     0037     .039     .036    .032     .024      3.38     1.64    7

Cortlcosterone     .013     .065     .139     .070    .050     .039      8.78     5.35    6

17-OH DOC          .041     .101     .084     .076    .062     .055      8.41     4.28    5

Desoxycortlcos-
 terone            .197     .313     .143     .069    .035     .020     12.55     8.53    6

17-OH Proges-
 terone            .042     .120     .213     .211    .078     .031     14.76    11.35    7
Continued on following page
                                               -16-

-------
Table 3 Continued -
Topical Administration*—
Steroid
Progesterone
Hud not one
acetonlde
Dexamethasone
Estradlol
Testosterone
Testosterone
acetate
Testosterone
proplonate
Dehydroeplan-
drosterone
Androstenedlone

Tirtie (hrs)
0-12
.208
.002
.005
.008
.147
.103 '
.061
.265
.183

Absorption
12-24 24.48
.264
.011
.003
.056
.364
.133
.096
.446
.334
.135
.012
.004
.099
.156
.048
.035
.249
.155
Rate (X/hr)
48-72
.045
.016
.003
.101
.066
.015
.015
*
.091
.076
72-96
.024
.008
.002
.107
.036
.007
.009
.046
.043
96-120
.011
.005
.002
.103
.018
.004
.005
.028
.028
Total
Absorption
% of dose
10.81
1.34
.40
10.62
13.24
4.62
3.44
18.45
13.47
S.T).
5.78
1.05
.23
4.86
3.04
2.28
1.03
7.71
5.56
No.
Subjects
6
9
3
3
17
6
9
6
11
From Feldmann and Haibach.  1969
•Corrected for l.v.  recovery
                                                -17-

-------
average amount of absorption for this series of steroids is roughly 10 percent
with a fifty-fold difference between the least and the most absorbed.   However,
inspection of the standard deviation column (S.D.) shows that there is a great
deal of uncertainty in this methodology with penetran-;s like hydrocortisone
having a standard deviation of ^ 85 percent.  The authors point out a  physical
relationship between the number of hydroxyl  groups on  the steroids  and their
total absorption; corticosterone and 17-OH desoxycorticosterone each have two
while desoxycorticosterone and 17-OH progesterone each have one.   For  each pair,
the absorptions are reasonably close.  Also, the series of steroids shows a
stepwise increase in absorption for each hydroxyl  removed (Feldmann and Maibach,
1969).
     Table 4 presegts further in vivo percutaneous absorption studies  by Feldmann
and Maibach (1970) on some organic compounds.   The values for percent  recovery
shown in the first part of the table indicate that several  of the compounds are
poorly absorbed and excreted (i.e., the correction factor for hexachlorophene
is greater than 20 which means that the percent recovery for this compound in
urine after topical  administration has  to be multiplied by  this factor).  Three
compounds were deemed too toxic to test intravenously  in man and thus  were
administered to guinea pigs.  Nicotinic acid,  nicotinamide, hippuric acid and
phenol were assumed to be excreted similarly to salicylic acid, while  thiourea
was assumed to behave like urea.  The second part of Table 4 gives  the absorption
after topical administration.  Three compounds, caffeine, dinitrochlorobenzene,
and benzoic acid, had total  absorptions of about 50 percent after being corrected
for incomplete absorption following i.v. administration.  Three other compounds,
nicotinic acid, hippuric acid, and thiourea displayed  total absorptions of less
                                      -18-

-------
       TABLE 4.   PERCUTANEOUS ABSORPTION OF SOME ORGANIC  COMPOUNDS  IN MAN



I.V. Administration—
Compound
w-
Caffeine
Chloramphenlcol
Colchlclne
Dlethyltoluamlde
D1n1trochlorobenzene
Hexachlorophene
Nitrobenzene
Potassium thlocyanate
Salicylic add
Urea
Butter Yellow*
Malathlon*
Methyl chol anthrene*
X Recovered
59.4
67.4
27.9
52.3
64.0
4.4^
60.5
10.2
89.8
71.7 .
58.6
76.0
18.2
Half Life (hrs)
6 r
6
4
4
4
48
20
12
4
8
6
4
14
•Recovery In urine  after  l.v. administration In guinea  pigs.
                                      -19-

-------
Table 4 continued—
Topical  Administration*—
Compound
Absorption Rate
ITjme (nrsj
1 0-12
Acetyl salicylic
•eld
Benzole add
Butter yellow
Caffeine
Chloramphenlcol
Colchlclne
Dlnltrochloro-
benzene
Dlethyltoluamlde
Hexachlorophene
Hlppurlc acid
Malathlon
Methyl chol anthrene
Nicotfnlc add
Nlcotlnamlde
Nitrobenzene
Pa raaminoben zoic-
acid
Phenol
Potassium
thiocyanate
Salicylic add
Thlourea
Urea
.141
3.036
.215
.559
.007
.036
3.450
.773
.029
.005
.313
.062
.000
.019
.022
.159
.254.
.051
.116
.046
.008
12-24
.438
.340
.685
1.384
.019
.038
.565
.331
.031
.003
.170
.329
.002
.168
.022
.648
.091
.060
.535
.035
.021
24.48
.334
.055
.289
.855
.021
.033
.134
.084
.020
.001
.044
.258
.001
.177
.013
.444
.010
.078
.356
.010
.051
(Wose/hr)
48-72
.147
.000
.083
.109
.022
.040
.045
.036
.028
.001
.017
.127
.001
.088
.013
.196
.601
.097
.156
.008
.073
72-96
.076
.000
.054
.032
.015
.025
.018
.016
.034
.001
.011
.064
.002
.052
.011
.058
.000
.100
.080
.007
.075
96-120
.060 .
.000
.022
.014
.012
.004
.009
.012
.030
.001
.006
.045
.007
.031
.006
.044
.000
.093
.033
.007
.034
Total
Absorption
S of dose
21.81
42.62
21.57
47.56
2.04
3.69
53.14
16.71
3.10
.21
7.84
16.81
.34
11.08
1.53
28.37
. 4.40
10.15
22.78
.88
5.99
S.O
No.
Subjects
r
3.11 S
16.45 <
4.88 4
20.99 12
2.46 6
2.50 6
12.41
4
5.10 4
1.09 7
.09 7
2.71
5.16
7
3
.09 3
6.17 7
.84 6
2.43 13
2.43
3
6.60 6
13.25 17
.22 3
1.91
4
From Feldmann and Maibach, 1970
•Corrected for  l.vf recovery.
                                             -20-

-------
than 1 percent of the applied dose.   The range for total  absorption  for  these
compounds is greater than 250-fold,  much larger  than  the  series  of steroids  in
Table 3.  However, as with the steroids, the standard deviations are very  high
for the compounds, indicating a large degree of  uncertainty,  with chloramphenicol
showing a standard deviation of greater than 100 percent.   The authors point
out two examples of closely related  compounds showing great differences  in
penetration:  benzoic acid was absorbed 200 times more than  its  glycine  conjugate,
hippuric acid; nicotinic acid showed minimal  penetration  while 10 percent  of
its amide, nicotinamide, was absorbed.  The authors also  generally found a
good correlation between the maximum penetration rate for  each compound  and
its total absorption.
     Table 5 presents similar in vivo absorption studies  by  Feldmann and Maibach
(1974) on some pesticides and herbicides.   The authors used the  same experimental
method as discussed previously to study 5 organophosphates,  3 chlorinated
hydrocarbons, 2 carbamates, and 2 herbicides. The total  excretion from  the
i.v. dose varied over a wide range with dieldrin, aldrin,  and carbaryl at  3-7
percent and malathion, baygon, and 2,4-D being over 80 percent.   For the topical
administration, diquat was the only  compound with only slight penetration;
carbaryl, on the other hand, was nearly completely absorbed after multiplying
the topical results by the large correction factor obtained  from the i.v.
administration.  The authors discuss the rather  large standard deviations  of
1/3 to 1/2 of the rcson value found in their studies.   They claim that the
difference is due to actual differences between  people rather than experimental
error as repeat studies on the same  subject gave similar  results.  The authors
assume a normal distribution and find that 1 person in 10 will  absorb twice
                                      -21-

-------
                                      TABLE S.  PERCUTANEOUS PENETRATION OF SOME PESTICIDES AND HERBICIDES IN HAN
               I.V.  Administration—
i
ro
                                            Absorption rate (X dose/hr)(t1me period In hr)
Total Absorption
Compound
Azodrln
Ethlon
Gut hi on
Ma lath Ion
Parathlon
Baygon
Carbaryl
Aldrln
Dleldrln
Llndane
2.4-D
Dlquat
Time (hrs)
^ 0-4 4-8
1.816
0.832
1.513
12.949
0.035
10.361
0.459
0.224
0.038
0.688
3.001
9.328
2.721
1.041
1.204
5.571
1.321
7.290
0.394
0.091
0.067
0.611
4.063
1.544
8-12
1.701
1.892
1.590
2.420
2.508
1.478
0.211
0.113
0.074
0.552
5.312
1.825
12-24
1.000
0.791
1.041
0.368
1.124
0.192
0.102
0.040
0.046
0.244
1.728
0.292
24-48
0.679
0.316
0.813
0.052
0.469
0.064
0.037
0.023
0.046
0.232
0.737
0.127
48-72
0.341
0.123
0.458
0.017
0.135
0.053
0.021
0.013
0.013
0.132
0.275
0.059
72-96
0.173
0.071
0.257
0.008
0.069
0.047
0.011
0.011
0.015
0.125
0.153
0.054
96-120
0.088
0.065
0.127
0.004
0.037
0.043
0.008
0.008
0.008
0.102
0.097
0.045
% Dose
67.7
38.4
69.5
90.2
45.8
83.8
7.4
3.6
3.3
24.6
100.0
61.2
Half-life
SO flir)
5.3
3.6
6.9
9.7
5.3
7.2
2.2
0.9
1.0
6.1
2.5
16.0
20
14
30
3
8
8
9
6
28
26
13
4
Topical Admlnlstratlon--
Azodrln
Ethlon
Guthlon
Halathlon
Parathlon
Baygon
Carbaryl
Aldrln
Dleldrln
Llndane
2.4-D
Dlquat
0.057
0.004
0.044
0.089
0.007
0.351
0.005
0.086
0.143
0.064
0.009
0.005
0.048
0.005
0.202
0.408
0.116
0.705
1.212
0.074
0.137
0.135
0.012
0.002
0.092
0.026
0.294
0.396
0.243
1.204
3.027
0.078
0.135
0.245
0.020
0.005
0.121
0.036
0.276
0.149
0.202
0.543 -
1.944
0.079
0.093
0.113
0.029
0.003
0.183
0.044
0.207
0.029
0.110
0.093
0.853
0.066
0.066
0.088
0.068
0.003
0.147
0.041
0.125
0.008
0.062
0.023
0.277
0.053
0.060
0.066
0.082
0.003
0.113
0.015
0.059
0.006
0.036
0.020
0.154
0.060
0.043
0.051
0.043
0.002
0.073
0.011
0.040
0.005
0.029
0.028
0.105
0.061
0.034
0.048
0.027
0.001
14.7
3.3
15.9
8.2
9.7
19.6
73.9
7.8
7.7
9.3
5.8
0.3
7.1
1.1
7.9
2.7
5.9
• " 5.2
21.0
2.9
3.2
3.7
2.4
0.1




               From Feldmann and Malbach, 1974
               •Corrected for l.v. recovery.  There were 6 subjects for each compound for both l.v.  and topical  administration.

-------
the mean value while 1 in 20 will  absorb 3 times this amount.   They have also
found that experimental  subjects differ by a  factor of 5 in  the amount  of
percutaneous absorption.  The authors also comment that the  experimental  subjects
did not sweat extensively such as  field worker might when exposed  to these
pesticides; the effect.of sweating should be  studied.  Also, Larry Hall  (personal
communication) points out that the skin penetration tables in  Feldmann  and
Maibach's studies are accurate only if the rate of elimination is  much  greater than
the rate of skin absorption.

5.2 IN VIVO ANIMAL STUDIES
     The toxicity of many penetrants prohibits in vivo human studies, thus
researchers have had to  use animal  models to  obtain absorption data. Unfortunately
there are a number of problems associated with the extrapolation of animal  data
to humans.  Animal species variation, different sites of application, shaved
skin vs. unshaved, occluding or restraining devices, and skin  metabolism (or
lack thereof) are some of the factors which need to be considered  when  using
animal studies to predict human absorption.
     Bartek et al. (1972) compared percutaneous absorption in  rats, rabbits,  minia-
ture swine, and man;  Table 6 shows the results of their study.  Radiolabeled
compounds were applied to the shaved skin of  the back and protected by  a non-
occluding device.  In general, the penetration through the skin of the  pigs
and man was similar and  much slower than it was through rat  and rabbit  skin.
For the compounds studied, haloprogin was completely absorbed  by the rat  and
rabbit, while only 11 percent was absorbed by man.  About half of  the caffeine
applied was absorbed by  all  of the animals plus man.  The results  of this
study show that absorption in the rabbit and  rat would not be  predictive of that
in man while the miniature swine appears closest to man.

                                      -23-

-------
     TABLE 6.   IN VIVO  PERCUTANEOUS ABSORPTION BY RAT. RABBIT, PIG. AND MAN
Penetrant
Haloprogln
N-Acetylcycte1ne
Testosterone
Cortisone
Caffeine
Butter Yellow
Percent
Rat
95.8
3.50
47.4
24.7
53.1
48.2
dose absorbed*
Rabbit
113.0
1.98
69.6
30.3
69.2
100.0
Pig
19.7
6.00
29.4
4.06
32.4
41.9
-Nan
11.0
r
2.43
13.2"
3.38"
47.6"
21.6"
From Bartek,  laBudde,  and Malbach. 1972.
•Corrected for  recovery following l.v. administration
"Human data  taken from Feldmann and Malbach, 1969 and 1970.
                                        -24-

-------
     Bartek and LaBudde (1975) also studied the absorption of 4 pesticides in
the rabbit, pig and squirrel  monkey using the same techniques as their  previous
study.  The results were compared to man where the site of application  was the
forearm (Table 7).  The in vivo absorption of pesticides in the rabbit  was
greater than in man, while absorption in the pig and squirrel monkey was  closer
to that in man.
     Shah^^l_. (1981) investigated the dermal penetration of some insecticides
in mice.  Acetone solutions of the radiolabeled test penetrants were applied
to shaved backs of female mice.  Mice were placed in metabolism cages equipped
with C02 trapping devices and urine collection jars.  Urine, blood, specific
tissues, and organs were sampled at various time intervals.  The remaining
portion of the body was termed carcass.   Table 8 shows the percent  of radio-
labeled penetrant recovered in various fractions at 5, 15, and 60 minutes
following topical administration.  Recovery of radioactivity was 90% or more
in all cases.  The authors also tried to relate physical properties such  as
molecular weight, solubility, and partition coefficients to dermal  absorption.
Partition coefficients for chloroform-water, olive oil-water, and benzene-water
were determined using radiolabeled compounds and a high-speed centrifuge.
Table 9 presents a comparison between physical parameters and rates of  penetra-
tion.  The authors did not draw any conclusions from their study other  than
that there is a lack of correlation among partition systems themselves.  One
interesting aspect of this study is that an aqueous wettable powder formulation
of carbaryl penetrated more rapidly than did the acetone formulation.
     Researchers have attempted to bridge in vivo animal and human  dermal
absorption studies by transplanting human skin to the athymic (nude) mouse.

                                      -25-

-------
TABLE 7.  IN VIVO PERCUTANEOUS ABSORPTION OF SEVERAL PESTICIDES BY RABBIT,  PIG,
                            SQUIRREL  MONKEY, AND HAN
Percent Dose absorbed*
Pesticide

DDT
Llndane
Parathlon
Malathion

Rabbit
46.3
51,2
97.5
64.6

Pig
43.4
37.6
14.5
15.5

Monkey
1.5
16.0
30.3
19.3

Man**
10.4
9.3
9.7
8.2
From Bartek and LaBudde, 1975

•Corrected for recovery following 1,v. administration (except  the monkey data).
**Human data from Feldmann and Malbach, 1974
                                      -26-

-------
                       TABLE 8.  GEOMETRIC MEANS OF PERCENTAGE  14C RECOVERED  IN  VARIOUS  FRACTIONS AT 5, 15.
                                                   AND 60 MIN POST APPLICATION
Toxl cant
Carbamates
Carbaryl
Methomyl
Carbofuran
Organophosphates
Parathlon
Malathlon
Chlorpyrlfos (methyl)
Chlorpyrlfos
Botanical type
Nicotine
Permethrln
Chlorinated hydrocarbons
DDT
Hexachloroblphenylb
4-Chloroblphenyl
Chlordecone
Dleldrln
Blood
5 15
mln mln

0.1 0.3
0.1 1.8
<0.1 1.7

<0.1 0.2
0.3 0.9
0.3 0.7
<0.1 0.4

0.6 2.1
<0.1 1.0

<0.1 <0.1

-------
                  TABLE  9.  COMPARISON BETWEEN PHYSICAL PARAMETERS AND GEOMETRIC MEANS OF  RATE  OF PENETRATIONS
ro
oo
Partition Coefficients
Molecular ^2°
Toxicant Weight Solubility
Carbamates
Carbaryl
Met homy 1
Carbofuran
Organophosphates

203
162
221

Parathlon 292
Malathlon 330
Chlorpyrlfos (methyl )318
Chlorpyrlfos 350
Botanical type
Nicotine
Permethrln

162
390

40 ppm
58.000 ppm
700 ppm

24 ppm
145 ppm
4.7 ppm
2 ppm

Mlsdble
0.07 ppm
CHC13
Water

277
15
32

433
112
92
374

2
269
Olive Oil-
Water

46
0.1
5

1738
56
245
1044

0.02
360
Benzene- TO. 8
Water (mln)

139
40

659
91
116
480

0.4
80

12.8+4.1
13.3+2.8
7.7+1.5

66.0+21.9
129.7+47.7
51.6+6.5
20.6T5.9

18.2+2.4
5.9+1.3
of
1
mln

22.2
18.8
28.5

4.4
5.5
8.3
15.7

5.2
36.2
Geometric
percentage
5
mln

31.5
24.7
32.6

9.8
13.4
14.3
28.8

27.9
40.9
15
mln

56.0
55.5
71.7

8.3
22.7
32.2
64.4

59.5
63.1
means
penetration
60 480 2880
mln mln mln

71.7 88.5
84.5 88.3
76.1 94.7

31.9 85.4
24.6 66.7
54.4 78.2
69.0 73.9

71.5 90.7
79.7 88.1

98.9

99.6
97.8

	
Chlorinated hydrocarbons
DOT
Hexachl orobl pheny 1 a
4-Ch1orob1pheny1
Chlordecone
Dleldrfn
355
361
188
490
384
1.2 ppb
0.953 ppb
<0.6 ppm
0.4 ppm
0.18 ppm
532
2465
684
190
1775
887
482
185
282
170
1398
144
105.4+25.6
43.8+7.0
16.8T2.3
41.3T8.8
71.7+17.3
3.9
17.3
6.5
18.8
1.2
12.3
28.7
13.5
34.4
3.1
21.7
44.7
53.5
42.5
26.1
34.1 71.1
55.3 66.8
84.5 97.5
54.0 65.9
33.7 82.6
91.3
94.0
    From Shah, et al.  1981  (and  references therein)

-------
Krueger and Shelby (1981)  used human  skin  obtained  from skin grafts or organ
donors to graft to nude mice.   The authors found  that the  human  skin grafts
undergo a proliferative response when 10  ng of the  tumor promoter 12-0-
tetradecanoyl  phorbol  13-acetate is applied while nude mice do not  respond to
this dose.  The authors state  that their  studies  show that the unit function of
human skin after transplant is similar to that prior to transplant.

5.3  IN VITRO
     In vitro studies  have also been  used extensively to estimate absorption.
For these studies, a piece of  excised skin is attached to  a diffusion apparatus
which usually consists of  a top chamber to hold the applied dose of the penetrant
plus any solvent, an 0-ring to hold the skin in place, and a temperature con-
trolled bottom chamber containing saline  or other solvents plus  a sampling
port to withdraw fractions for analysis (Fig. 2).  Human forearm skin is difficult
to obtain, thus it is  common practice to  use abdominal skin collected at autopsy.
For most studies, the stratum  corneum is  heat separated from the epidermis and
dermis, then studied by itself.
     Franz (1975) studied  the  in vitro permeability of 12  organic compounds which
had been previously studied in vivo in man. Special emphasis was given to
duplicate the in vivo conditions, such as amount  of dose applied, in order to
show how accurately in vitro absorption studies can reflect the  living state.
Each piece of skin was mounted in a diffusion cell  (diffusion area of either
1 cm^ or 2.5 cm^) with the epidermal  side of the  skin exposed to ambient air
while the dermal side was  bathed in a saline solution containing an antibacterial/
antifungal plus buffers.  The  temperature was maintained at 37°C by a water
                                      -29-

-------
                                                         Donor
    Skin
"0" Ring
    Temperature
    Jacket
                                                                              Sampling Port
                                                                            Outflow
                                                                            Inflow
                                                         Stirring Bar
            Figure 2. Schematic representation of a diffusion cell with top
                     open to the ambient environment.(Franz, 1975)
                                           -30-

-------
jacket which surrounded the chamber.   A small  amount of the radiolabeled penetrant
in the same dose range as the in vivo studies  was  dissolved in  acetone and
spread across the entire exposed surface with  the  acetone evaporating in less
than one minute.  At selected intervals after  the  addition of penetrant to the
epidermis the dermal bathing solution was removed  in its entirety,  gelled and
analyzed in a liquid scintillation spectometer.  Either the total absorption
was determined by one sample taken at 24-hr  intervals or the kinetics of the
absorption process were determined by taking frequent samples throughout the
day.
     Table 10 shows the total  absorption of  12 organic chemicals that have been
studied both in vivo and in vitro.  Highly water-insoluble compounds were not
selected for in vitro study since their permeability might be limited due to
insolubility in the dermal bathing (saline)  solution.  In viewing the data from
Table 10 only two compounds, chloramphenicol and benzoic acid appear to be in
quantitative agreement.  The in vivo studies are from Feldmann  and  Maibach and,
as discussed previously, are from 5-day urine  collection of an  applied dose to
the forearm.  The in vitro studies, on the other hand, use human adominal
tissue, last for two days, and have a much smaller receptor volume  when compared
to the entire body (i.e., you might see a build-up of penetrant concentration
in the receptor cell that you would not see  when the dose is applied in vivo).
     Franz (1978) further investigated the apparent differences between the jji
vivo and in vitro approaches.  Since in only two of the twelve  compounds studied
was there no radiolabeled compound in the urine on day 5, the author thought
that this collection period might be inadequate and lead to underestimation of
                                      -31-

-------
         TABLE 10.  COMPARISON OF HUMAN IN VIVO AND W VITRO ABSORPTION


                       Total Absorption (expressed as percent of applied dose)
Compound
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Hlppurlc add
Nlcotlnlc acid
Thlourea
Chloramphenicol
Phenol
Urea
Nlcotlnamlde
Acetyl salicylic add
Salicylic acid
Benzole add
Caffeine
D1 n1 1 rochl orobenzene
In V1vo«
0.2 + 0.1 [7]
0.3 + 0.1 [3]
0.9 l 0.2 [3]
2.0 + 2.5 [6] .
4.4 + 2.4 [3]
6.0 +_ 1.9 [4]
11.1 +6.2 [7]
21.8 +_ 3.1 [3]
22.8 + 13.2 [17]
42.6 +_ 16.5 [6]
47.6 ^ 21.0 [12]
53.1 + 12.4 [4]
In V1trob
1.2 (0.8. 2.7) [15]
3.3 (0.7, 8.3) [19]
3.4 (2.4. 5.5) [52]
2.9 (1.0. 5.7) [12]
10.9 (7.7, 26) [7]
11.1 (5.2. 29) [22]
28.8 (16, 65) [21]
40.5 (17, 49) [14]
12.0 (2.3, 23) [10]
44.9 (29, 53) [18]
9.0 (5.5, 20) [17]
27.5 (19. 33) [18]
From Franz, 1975 (In vivo data from Feldroann and Malbach.  1970)

'Mean + standard deviation.  The figure 1n brackets Is the number of subjects
studied*.
"Median with 95$ confidence Interval  given 1n parentheses.
                                      -32-

-------
the actual  amount absorbed.   Also,  there  are  differences in the permeability
of skin from different sites of the body.  Franz  restudied four compounds
(hippuric acid, nicotinic acid, thiourea, and caffeine) for which there was
poor agreement between the in vivo  and  in vitro results.  This time adominal
skin was used for both the in vivo  and  in vitro tests  plus urine was collected
in the in vivo tests until background levels  of radioactivity were reached.
Also, to minimize the differences that  might  be caused by desquamation of the
epidermis in vivo, the surface of the skin was washed  with acetone and water
24 hours after application of the test  compounds  in both sets of experiments.
     Table 11 shows the comparison  between human  in vivo and in vitro absorption,
using Franz's refined techniques, for the four compounds originally showing
poor agreement.  Hippuric acid, thiourea, and caffeine gave excellent quantitative
agreement, but nicotinic acid displayed the lack  of agreement found in the
previous study (see Table 10).  Franz found that  less  than 15 percent of an
intravenously administered dose of  ^C-nicotinic  acid  was excreted in the urine;
Feldmann and Maibach (1970)  did not directly  measure the excretion of nicotinic
acid following an i.v. dose, but assumed  on the basis  of similar chemical
structures that it would behave like salicylic acid, a compound in which 90
percent of the i.v. dose is excreted.  Thus,  when the  value of 0.32 percent
absorption observed in vivo in humans is  corrected for incomplete urinary
excretion, the value of 2.1 percent is  in better  agreement with the value
observed in vitro.  Franz (1978) states that  based on  the data obtained from the
compounds studied to date, it appears that in vitro studies accurately portray
the phenomenon of absorption as it  occurs in  living man.
                                      -33-

-------
         TABLE 11.  COMPARISON OF HUMAN IN VIVO AND IN VITRO ABSORPTION
                            USING REFINED TECHNIQUES

                    Total  Absorption (expressed as  percent  of applied  dose)
Compound
1. N1cot1n1c Acid
2. Hlppurlc Add
3. TMourea
4. Caffeine
In V1voa
0^32 +. 0.10 [3]
1«0 + 0.4 [6]
3.7 + 1.3 [4]
22.1 +. 15.8 [4]
Tb
21
3
21
7
In V1troa
2.3 _+ 0.9 [4]
1.25^0.5 [4]
4.6 + 2.3 [5]
24.1 +; 7.8 [4]
From Franz, 1978
aHean ± SO.  The figure 1n brackets Is the number of subjects  studied.
bNumber of days urine was collected.
                                      -34-

-------
     Bronaugh et al. (1982)  compared in vivo and  in  vitro  absorption through
female rat skin using benzoic acid,  acetylsalicyclic acid, urea,  and caffeine.
For the in vivo studies, petrolatum  was used as a nonvolatile  vehicle  because
of its ability to adhere to  the skin; in this manner the concentration was
known and, therefore, permeability constants could be calculated.   Full thick-
ness, lightly shaved skin (with the  subcutaneous  fat removed)  was  used in a
standard diffusion cell  for  the in vitro experiments.  Since the  permeability
                                                                         n
constant (kp) is defined as  the steady-state rate of absorption  (amount/cm /hr)
                                                                    o
divided by the concentration of solute applied to the skin (amount/cm  ), a kp
value (cm/hr) is obtained by the diffusion cell procedure. The  determination
of a kp value from in vivo data is based on the measurement of absorption rate.
This rate is estimated by considering the absorbed compound to accumulate in
both the body of the animal  and the  excreted waste.   The kp was  calculated from
the rate of body accumulation plus the rate of excretion divided  by the
concentration of solute in vehicle.   The quantitative agreement  between the i_n_
vivo and in vitro work of Bronaugh et al. (1982)  appears good  as  shown in
Tables 12 a, b, and c.
                                      -35-

-------
      TABLE 12a.  PERCUTANEOUS ABSORPTION OF ACETYLSALICYLIC  ACID  IN  RATS
                                   Absorption (X of applied dose)3
Days	In V1vo	In Vitro
1
2
3
4
5
Total
8.5 i 1.6
7.9 +. 2.0
4.0 + 0.9
2.8\+ 0.5
1.9 + 0.5
24.8 + 4.4
8.8 +_ 1.2
8.5+^ 1.2
4.6+0.5
4.3^0.4
2.9 + 0.1
29.0 _+ 0.1
From Bronaugn et al. 1982
aResults are expressed as the Y+_ SE of four or five determinations.
  TABLE 12b.  Iti VIVO VS. IN VITRO PERCUTANEOUS ABSORPTION THROUGH RAT SKIN8
                             Rate (ng/hr/cm^)   Permeability  constant
Test Compound
Caffeine
Acetyl salicylic add
Body
16.3
0
Urine
27.8
11.4
In V1vo In Vitro
2.1 x 10-4(7) 3.1 x 10-4(6)
5.2 x 10'5(7) 6.5 x 10-5(5)
From Bronaugh et al. 1982
'Compounds In a petrolatum vehicle were applied to a  2.0-cm^  area  of  skin  on the
 living animals and  1n diffusion cells.  Results are  the means  of  the number of
 determinations In parenthesis.

    TABLE 12c.  COMPARISON OF PERMEATION VALUES WITH  THOSE  OF OTHER STUDIES
                                 In V1vo                        In Vitro
Test Compound
Benzole add
Acetyl salicylic acid
Urea
Rat
(Petrolatum)
37.1
24.8
8.1
Human0
(Acetone)
42.6
21.8
6.0
Rat
(Petrolatum)
49.1
29.0
7.2
Human"
(Acetone)
44.9
40.5
11.1
From Bronaugh et al. 1982
aValues from Feldmann and Mai bach (1970)
bValues from Franz (1975)
                                      -36-

-------
             CHAPTER 6 - THEORETICAL TREATMENT OF DERMAL  ABSORPTION

6.1  PICK'S LAW APPLIED TO DERMAL ABSORPTION
     For those unfamiliar with Pick's Laws of Diffusion,  it  may  be helpful
to reference a basic physical  chemistry text such as Moore,  1962.  Scheuplein
and Blank (1971) developed an  integrated form of Pick's law  to express  diffusion
through the skin barrier.  The flow across the membrane is called the flux,
Js, and the expression for the steady-state flux of solute across an inert
membrane is given by:
                   Js  =   D(d - C?)
                                                               O    1
        where    Js  =  steady-state flux of solute  (moles  cm"'1 hr"1)
                 D  =  Average membrane diffusion coefficient  for solute
                         o    1
                      (cnrsec"1) (sometimes interchanged with  Dm to represent
                                  diffusion coefficient through skin membrane)
                C  =  Concentration of solute
                6  =  Membrane thickness (cm)
For skin, the stratum corneum is not an inert structure but  one with an affinity
for the applied solute; thus, the concentrations at the surfaces of the membrane
are hot usually equal to the concentrations in the external  solutions.   The
correlation between external and surface concentrations can  be stated in
terms of the sol vent -membrane distribution coefficient, Km.  The intergrated
form of Pick's law then becomes Js  =  KmDAC<;
                                          6
                     and        kp  =  Km D
                                        6
          where    ACS  =  Concentration difference of solute across  membrane
                           (moles cm~3)
                                      -37-

-------
                    k. =  Permeability constant  for  solute  (cm  hr~)

                                 Js  =  kp ACS
     In the application of the expanded form of  Pick's  law,  i.e.,  Js  =  kp&Cs  =
      , only the skin membrane thickness,  diffusivity,  and the  partition
  6
coefficient for the solute-solvent  membrane  (Km)  are  considered.   This equation
describes reasonably well  the permeability of the skin  to  non-electrolytes,  but
this simplified expression applies  only to steady-state permeability.  The time
it takes for the permeability to reach a steady-state is called the  lag  time
(T) for diffusion.  For an simple,  ideal  membrane, the  lag time is related to
the diffusion constant by  T = «2  (Dugard, 1983).  Another factor  of interest  is
                              6Dm
the amount of penetrant remaining dissolved  within the  stratum corneum at the
end of a short period of contact.  The penetrant  that is within the  stratum
corneum cannot be quickly  washed away and eventually  will  enter the  body (unless
volatile or lost through desquamation), leading to the  term "reservoir effect."
A diagram showing the concentration profile  across stratum corneum during steady-
state absorption, assuming uniform  properties across  the membrane, is shown  in
Figure 3.
     A corollary of Pick's law is that the chemical potential  of a penetrant is
maximal in a saturated solution and therefore a maximum absorption rate  occurs
from a saturated solution.  The chemical  potential  of a particular chemical  is
the same in all saturated  solutions (i.e., neat liquid  or  solid),  regardless of
solvent.  This means that  there is  a single  maximum absorption rate  definable
for any penetrant.  The absorption  rate for  an ideal  system is proportional  to
the degree of saturation,  and, thus, all  half-saturated solutions  of a penetrant
have equal chemical potential and all give half the maximum possible absorption
                                      -38-

-------
                   Penetrant

                   Concentration
i
to
vo
I
                                                                                                           c=o
                                   Outer
                                  Solution
Membrane
  Inner
Solution
                                         Figure  3. Diagram showing the concentration profile across

                                                   stratum corneum during steady-state absorption

                                                  (From Dugard, 1983)

-------
rate (Dugard, 1983 and  ref.  therein).
     Knowledge of any physical equilibrium condition may be used to relate the
chemical  potential in one vehicle  or physical  state to that in another.
Predictive treatments based  on solubilities or on equilibria are largely untested
and depend on several systems behaving close to the ideal, thus any predictions
should be regarded as qualitative.  Reasons for the breakdown of solubility-
derived predictions include  solvent damage to  stratum corneum, deviations from
ideality near penetrant saturation in vehicles, variations in stratum corneum
hydration in contact with different solvents,  and entry of vehicle components
into the stratum corneum to  alter  its solvent  properties (Dugard, 1983).
6.2  NON-STEADY STATE DIFFUSION    '
     The diffusion of the penetrant across the stratum corneum is usually the
rate determining step in the dermal absorption process; however, there are
instances where this is not  the  case.  Higuchi (1962) has considered the
condition where the penetrant is entirely dissolved in the vehicle and its
diffusion in this medium is  slow.  Assuming that any penetrant reaching the
stratum corneum is immediately absorbed  and that not more than 30% of the
original  amount of penetrant is  absorbed, then the total amount of penetrant
(Qt) released from the  vehicle by  time t is approximately
                              Qt  = 2c AL
where Dv is the diffusion constant  of the  penetrant and Cv is the concentration
of the penetrant in the vehicle.
     A second situation where  diffusion  across the stratum corneum is not rate
limiting occurs when the vehicle  contains  suspended penetrant whose dissolution
is rate limiting.  Higuchi  (1960) described  the  release and absorption of the
                                      -40-

-------
penetrant when it is in small  particles  evenly  dispersed  in the  vehicle as
                      Qt = (2  Av * Sv> [.    A*   c  t ic 11/2
                                        1  +  Z  [AV-5V)/5VJ
where Av is the total amount of penetrant, dissolved and  suspended,  in the
vehicle per unit volume, and Sv is its solubility  in the  vehicle.  The rate
of absorption at a given time  t is obtained  by  differentiating Q^  (Dugard
1983, and references therein)  to yield
                 W = 1/2 ki^jsj1/2
If Av is much greater than Sv, these equations  reduce to
                  Qt = (2AvDvSyt)1/2 and dQ  ,/Av D  Sy\1/2

     Another type of non-steady state absorption can occur when  a  small amount
of penetrant is applied.  "Small" means  that the source of the penetrant is
significantly reduced in quantity by the absorption  process.  Figure 4 shows
the manner in which the absorption rate  changes with time for a  fast penetrant,
moderate penetrant, and a very slow penetrant,  when  depletion of the external
source of penetrant occurs (Dugard, 1983).
     Looking at Figure 4 we see that for curves A  and B a specific maximum
rate is achieved.  The time taken to reach the  maximum  rate, Tmax, is obtained
from the relationship Tmax = 62-h2  (Scheuplein and  Ross, 1974).  The thickness
                            ^
of the penetrant layer, h, is  usually small  in  comparison with 
-------
ro
i
                         o
                         o
                         vt
                         .Q
03
CC
                                                           Time
                              Figure 4. The pattern of changing absorption rate for small
                                       amounts of penetrant per unit area of skin. Curve A,
                                       moderately fast penetrant; curve B, slower penetrant;
                                       curve C, very slow penetrant. Curves A and B show the
                                       effect of the depletion of penetrant source. (Dugard, 1983).

-------
Ross (1974)'also state that the peak rate of absorption  is  directly  proportional
to the amount of penetrant applied per unit area,  which  they  designate  as  the
specific dose A.
     For very slow penetrants such as cortisone, the authors  found the  rate  of
absorption to be constant for a relatively long period of time  (such as curve C
in Figure 4).  This rate is approximately proportional to the specific  dose,
thus the flux can be expressed as:
                          J = k^A (slow penetrant, 6>h)
where k^ is a transfer coefficient.  Instead of permeability  constant (kp)
defined in terms of flux and concentration, the transfer coefficient (k^)
is defined in terms of flux and specific dose (A).  Unlike  the  permeability
constant, as A increases k^ gradually decreases because  the source of penetrant
comes to contain excess material  where additions cause no further increase in
absorption rate (Dugard, 1983).

6.3  CALCULATION OF kp and Km
     Scheuplein (1965) calculated kp values for the homologous  series of normal
primary alcohols Cj - CQ by measuring the in vitro permeability of the  alcohols
through human abdominal skin in a diffusion cell.   The sol vent-membrane distribution
coefficients Km, or partition coefficients, were computed from  the loss in
concentration of the original solution after equilibrium with the tissue,  i.e.:
     Km  =  moles alcohol absorbed per unit mass of dry  tissue
            moles alcohol in solution per unit mass of water
Penetration rates (Js) were obtained from the rate of increase  of concentration
on the receptor (bottom half of the diffusion cell).  Permeability constants kp
were computed directly from the linear portion of  the accumulation curve or  by
graphical procedures.
                                      -43-

-------
     Table.13 gives a summary of the  membrane  permeability  and  partition
coefficients for the normal  primary alcohol  series.  The  molecular  volume
increases by a factor of 4 within the series but  its effect on  the  rate of
diffusion is less than a factor of 1.6.   The Table also shows that  the permea-
bility constant increases as the molecular weight increases instead of
decreasing;  this is a result of the increasing alcohol-membrane solubility  as
a consequence of the decreasing polar character of the alcohols.
     It has  been known for some time  that the  permeability  of nonelectrolytes
through membranes increases  as the membrane  solubility of the penetrating
molecule increases.  More precisely it is the  solubility  of the penetrating
molecule within the membrane relative to  the solubility in  the  solvent (i.e.,
the membrane partition coefficient, Km) which  directly influences the permeability.
The actual partition coefficients of  lipophilic membranes are usually estimated
by approximating their lipophilic solubility with olive oil-water partition
coefficients.  Aqueous membrane partition coefficients Km have  been measured
for the alcohols and are compared with the olive-oil water  values K0 in Table
13.  It is apparent that there are large  differences between the two partition
coefficients.  The olive oil-water coefficient is a poor  approximation to Km
except in a narrow range near a value of  K0  =  Km  = 10.0.  The deviation near
the origin stems from the fact that Km must  be greater than 0.6, the lowest
conceivable weight fraction  of water  in the  tissue, while K0 values for water and
the very polar alcohols are in the order  of  10"^  owing to their low solubility in
olive oil.  From the deviation at higher  K values it is clear that  stratum
corneum is a less potent absorbent for strongly nonpolar  molecules  than is
olive oil (Scheuplein, 1965).
                                      -44-

-------
         TABLE 13.  MEMBRANE PERMEABILITY AND PARTITION COEFFICIENTS*

Water
Methanol
Ethanol
Propanol
Butanol
Pentanol
Hexanol
Heptanol
Octanol
If
Km
3.3
1.7
1.7
0.7
1.0
1.2
1.3
1.07
1.04
k V1/3
-£
8.65
5.85
6.64
2.95
4.51
5.72
6.50
5.55
5.60
«n
0.3
0.6
0.6
2.0
2.5
5.0
10.0
30.0
50.0
kp
1.0
1.0
1.0
1.4
2.5
6.0
13.0
32.0
52.0
"
18.02
32.04
46.07
60.09
74.12
88.15
102.2
116.2
130.2
V
18.02
40.05
59.3
74.9
91.5
107.9
124.8
141.3
157.4
"o
0.000
0.008
0.03
0.17
0.5
5.0
11.5
62.0
220.0
*  from Scheupleln, 1965
   kp  «  Permeability constant
   Km  «  Sol vent-membrane distribution coefficient
   K0  -  olive oil-partition coefficient
    V  •  Molecular volume
    M  -  Molecular weight
                                     -45-

-------
     Scheuplein et al. (1969)  also studied  a  series  of  steroids to see how
permeable these larger molecules,  some with polyfunctional  character, are
through skin.  Since steroid molecules have considerably  larger molecular
volumes than the linear primary alcohols  and  usually have several polar groups,
lower diffusion rates  are expected.  The  diffusion constants for the linear
primary alcohols are approximately the same with D = 10~9 cm^  sec"1, thus the
diffusion constants for the steroids  are  expected to be lower.  Similar
       s
experimental procedures (as for the linear  alcohol series)  were used to obtain
permeability constants for 14  common  steroids,  except that  partition coefficients
with amyl caproate and hexadecane  were also measured.
     The data in Table 14 were computed from  the steady state  portions of flux
vs. time curves; lag times (T) were extrapolated from the steady state portions
of these "penetrations curves" (see Fig.  5).  The concentration gradients used
in the experiments were extremely  small as  the  aqueous  donor solutions were
not saturated with steroid. The observed fluxes listed in  column 2, Table
15, therefore do not represent maximum obtainable fluxes  from  water solutions
of the steroids.  Since Pick's law is obeyed  at the  very  dilute concentrations
characteristic of saturated aqueous solutions of steroids,  maximum obtainable
fluxes may be computed.  These are listed in  column  3 of  Table 15 (Scheuplein
et al. 1969).
     Scheuplein et al. (1969)  chose the steroids for this study to include as
wide a range in polarity as possible  within a certain range of molecular weight.
Only a 25% difference  in molecular weight exists between  the lowest steroid
(estrone, MW = 270.3)  and the  highest (hydrocortisone,  MW = 360.4).  However,
there is approximately a 1000-fold difference in permeability  between these two
compounds.  The difference in  kp must lie in  the respective values for Km and
                                      -46-

-------
TABLE 14.  PERMEABILITY CONSTANTS AND PARTITION COEFFICIENTS FOR SOME STEROIDS
Steroid
Progesterone
Pregnenolone
Hydroxypregnenol one
Hydroxyprogesterone
Cortexone
Testosterone
Cortexolone
Cortlcosterone
Cortisone
Hydrocortlsone
Aldosterone
Estrone •
Estradfol
Estrlol
iCs
2.0
5.1
2.9
10.0
2.4
10
10
1.7
0.97
1.8
0.74
2.5
2.5
7.0
kp
1500
1500
600
600
450
400
75
60
10
3
3
3600
300
40
D
160
220
155
166
135
195
36.1
39.2
13.1
4.8
4.9
870
72.4
19.3
*m
104
50
43
40
37
23
23
17
8.5
7
6.8
46
46
23
*ac
56
52
49
46
30
16
11.2
6.8
1.52
1.30
	
80
66
1.64
Khex
17.0
4.2
1.6
2.5
3.0
2.6
0.1
0.024
0.28
0.009
	
3.0
0.63
0.23
From Scheupleln et al. (1969)
 AC  « Initial donor concentration 1n moles/cc x 10"9
  kn « Permeability constant 1n cm hr"1 x 10"6 (e.g. estrone kn • 3600 cm hr'1 x
   P   10-6j                                                  P
   D » Diffusion constant 1n crn^ see"* x lO"1^
  Km = Stratum corneum/water partition coefficient
 Kac * A"1*1 caproate/water partition coefficient
Knex > Hexadecane/water partition coefficient
           \
                                      -47-

-------

-------
                           TABLE 15.   STEROID  FLUXES


     The second and third columns are the observed  (J.exp) and the maximum
obtainable fluxes (J.max) with aqueous steroid solutions.  In the last column
are calculated values from 1n vivo penetration measurements where aqueous
solutions were not used.
Steroid

Progesterone
Pregnenolone
Hydroxypregnanol one
Hy d roxyp rogeste rone
Cortexone
Testosterone
Cortexolone
Corflcosterone
Cortisone
Hydrocorflsone
Aldosterone
Estrone
Estradlol
Estrlol
Js (exp)

30
51.3
17
60
10.9
40
7.5
1.04
0.097
0.055
0.022



J« (max) J« (In vivo

430
810
555
171
1350
349 150
[433]
287
65
23 7.0 - 20.0
[75]
173
13.2
58.4
From Scheuplein et al. (1969), and references  therein.

Js (exp) » Flux observed experimentally moles  cm^hr"1  x  10"13.

Js (max) • Predicted flux for saturated aqueous  steroid solutions.

Js (in vivo) = Values obtained from the literature.

[  ] » Indicates estimates of solubility used  for calculation.


                                      -49-

-------
Dm.  As shown in Table 14 the membrane-water partition  coefficient  Km  changes
only 15-fold through the group of steroids in comparison  to  the  approximately
1000-fold change in kn.  Thus, it is apparent that the  principal  determining
factor which decreases the permeability of the steroids is the decrease  in the
diffusion constant Dm.  Since the molecular volume of the steroids  is  about 3-4
times that of an average small molecule such as pentanol, we expect a  decrease
in Dm from this factor alone arising from the increased degree of chemical
interaction between the larger steroid molecule and the lipid-protein-H20 matrix
within the stratum corneum membrane.  Introducing additional  polar  groups into
the molecule lowers the diffusion constant still  further. Because  of  this
change in Dm within the series of steroids one cannot expect a proportionality
between permeability constant and partition coefficient.  Reference to Table 14
shows that there is no systematic relationship between  k« and Km values.
     Partition coefficients for the steroids between an ester, amyl caproate,
and water and between an alkane, hexadecane, and water  were  also measured.
Comparison of these data shows that the stratum corneum is much  more similar in
its solvating properties to partially polar amyl  caproate than to non-polar
hexadecane.  Although there doesn't appear to be any direct  proportionality
between Km and kp within this group of steroids,  there  does  exist the  possibility
of an useful correlation between the solubility of a steroid in  a particular
solvent and its permeability.  From the data in Table 14, Scheuplein et  al.
(1969) showed that one may expect a steroid with an amyl  caproate-water
distribution coefficient from 20-50 to have a permeability from  400-1000 x  10~6
cm hr'l.  The rest of the steroids may be broadly grouped as:
                                      -50-

-------
                                           kp x  IP"6  cm  hr"1
                   1-2                     1-10
                   2-10                   10-70
                  10 - 20                   70  -  400
                  20 - 50                  400  -  1000
A similar tabulation could be made for Khex and for Km but  the  latter  values
cannot be obtained as accurately and hexadecane does not  appear to  serve well
as an approximation for the solvating property  of  hydrated  stratum  corneum.

6.4  PREDICTION OF PERMEABILITY FOR SOME ALCOHOLS  AND STEROIDS
     Lien and Tong (1973) have tried to correlate  percutaneous  absorption data
for different types of drugs or organic compounds  with a  number of  physiochemical
constants by using a computerized multiple regression analysis  program.  By
comparing the equations of different sets of data  the authors sought to obtain
useful quantitative guidelines for predicting percutaneous  absorption.
     The authors took the absorption data and most of the chemical  constants
from the literature while octanol-water partition  coefficients  were either
experimentally determined or calculated.  The equation used in  the  computerized
                                                    o
regression analysis program is log BR = - kj (log  P)£ + ^2  log  P +  kj
(electronic term) + k4 (steric) + k$ where BR is the biological  response
expressed as the molar concentration (C) of the drug absorbed,  1/C  for a standard
response (such as erthema or vasoconstriction), or the permeability constant
(kp).  The coefficients kj through ks are obtained using  the method of
least squares.  For some cases the solubility in water appeared to  be  important;
therefore a log S term was also included in the analysis.  The  addition of
other steric or electronic terms, such as molar refraction, Taft's  polar sub-
                                      -51-

-------
stitute constant, and molecular weight,  significantly  improved the correlation
in some cases.
     For the alcohols and steroids,  the  variation  in permeability constants is
primarily due to the difference in lipophilic character.   By  comparing  an
equation derived from the permeability experiment  on aliphatic alcohols through
human epidermis with an equation from similar experiments  on  steroids,  i.e.,
(alcohols):  log kp = 0.934 log Km - 2.891  (n =  8, r = 0.986, s  = 0.121)
(steroids):  log kp = 2.626 log Km -7.537  (n  = 14, r = 0.931, s  = 0.377)
we see that the permeability of steroids through the epidermis is much  more
dependent on the partition coefficient into stratum corneum  (Km) as  compared
with the alcohols.  The lower intercept  of  the equation for  steroids as compared
to the equation for alcohols shows the stronger  hydrophobic  interactions between
the steroids and the epidermis than between the  alcohols and  the epidermis  (see
Fig. 6).  Tables 16 and 17 display the good correlations found for both alcohols
and steroids indicating that this approach  enables a chemist  to  predict the
relative degree of penetrant absorption  through  skin from  the physicochemical
constants of a series of compounds plus  the absorption data  of a few parent
molecules (Lien and Tong, 1973).
                                      -52-

-------
Figure 6.  Dependence of the permeability constant (log kp) on
          the stratum corneum/water partition coefficient (log Km)
          Equation A, log kp«0,934 log Km-2.891  is derived from
          the data of a alcohols absorbed through human epidermis;
          Equation B, log kp»2.626 log Km-7.537 is derived from
          the data of alcohols absorbed through human epidermis
          (From Lien and Tong, 1973).
                           -53-

-------
TABLE 16.  IN VITRO PERMEABILITY OF ALIPHATIC ALCOHOLS  THROUGH  HUMAN  EPIDERMIS*
Compound
Hater
Methanol
Ethanol
n-Propanol
n-Butanol
n-Pentanol
n-Hexanol
n-Heptanol
n-Octanol
•From Lien and
Log K0
• • •
-2.10
-1.52
-0.77
-0.30
0.70
1.06
1.79
2.34
long 1973
Log PO
• • •
-0.66a
-0.168
0.34b
0.84b
1.34b
1.84°
2.34b
2.84b

Log Km
-0.52
-0.22
-0.22
0.30
0.40
0.70
1.00
1.48
1.70

Obsd.
-3.00
-3.00
-3.00
-2.85
-2.60
-2.22
-1.89
-1.49
-1.28

Calcd.
• • •
-3.24
-2.97
-2.70
-2.43
-2.16
-1,88
-1.61
-1.34

D1ff.
• • •
0.24
-0.03
-0.15
-0.17
-0.06
-0.01
0.12
0.06

•Experimental determined value from Leo et al.,  1971.
Calculated value
P0 « Octanol/water partition coefficient.
K0 « Olive oil/water partition coefficient from  Scheupleln  (1965)
KID • Stratum corneum/water partition coefficient from Scheupleln  (1965)
kp » Permeability constant:  observed from Scheupleln (1965),  calculated  from
     log kp - 0.544 log P - 2.884     n - 8. r - .979. s -  0.150
                                      -54-

-------
     TABLE 17a.  IN VITRO PERMEABILITY OF STEROIDS THROUGH  HUMAN  EPIDERMIS*

Steroid                         Log Knex   Log Kac      Log Km        Log  Pe
Progesterone
Pregnenolone
Hydroxypregnenolone
Hydroxyprogesterone
Cortexone
Testosterone
Cortexolone
Cortlcosterone
Cortisone
Hydrocortlsone
Aldosterone
Estrone
Estradlol
Estrlol
1.23
0.62
0.20
0.40
0.48
0.42
-1.00
-1.62
-0.55
-2.05
. . .
0.48
-0.20
-0.64
1.75
1.72
1.69
1.66
1.48
1.20
1.05
0.83
0.18
0.11
. . .
1.90
1.82
0.21
2.02
1.70
1.63
1.60
1.59
1.36
1.36
1.23
0.93
0.85
0.83
1.66
1.66
1.36
2.78
(2.82)
(2.24)
(2.17)
1.72
1.94
(1.11)
0.66
0.15
0.20




•FromLlen and Tong, 1973
Knex « Hexadecane/water partition coefficient from Scheuplein et  al.,  1969
Kac » amyl caproate/water partition coefficient from Scheuplein et  al.,  1969
KID » stratum corneum/water partition coefficient from Scheuplein  et al., 1969
Pe * ether/water partition coefficient:   experimental  values without parentheses;
     values 1n parentheses were calculated from Flynn, 1971.
                                      -55-

-------
           TABLE 17b.   IN VITRO PERMEABILITY OF STEROIDS THROUGH HUMAN EPIDERMIS*
Steroid                      	Log K  (cm/hr)           Log D (cn^/sec x IP"13)
         	Obsd.°     Ca1td.b    D1ff.   "flbsd.3     Calcd.*-     D1ff.
Progesterone
Pregnenolone
Hydroxypregnenol one
Hydroxyprogesterone
Cortexone
Testosterone
Cortexol one
Cortlcosterone
Cortisone
Hydrocortlsone
A1 dosterone
Estrone
Estradlol
Estrlol
-2.82
-2.82
-3.22
-3.22
-3.35
-3.40
-4.12
-4.22
-5.00
-5.52
-5.52
-2.44
-3.52
-4.40
-2.23
-3.07
-3.26
-3.34
-3.36
-3.97
-3.97
-4.31
-5.10
-5.31
-5.36
-3.18
-3.18
-3.97
-0.59
0.25
0.04
0.12
0.01
0.57
-0.15
0.09
0.10
-0.21
-0.16
0.74
-0.34
-0.43
2.20
2.34
2.19
2.22
2.13
2.29
1.56
1.59
1.12
0.68




2.47
2.49
2.18
2.14
1.91
2.02
1.59
1.35
1.08
1.11




-0.27
-0.15
0.01
0.08
0.22
0.27
-0.03
0.24
0.04
-0.43




8 From Scheupleln et al.  1969
b Calculated from log kp  - 2.626  Kra - 7.537  (n - 14, r - 0.931. s - 0.377)
c Calculated from log D (cm2/sec  x  10'13) -  - 0.221 (log PJ2 + 1.170 log Pe + 0.734
  (n « 10, r - 0.961, s • 0.180)
                                            -56-

-------
             CHAPTER 7 - DERMAL ABSORPTION  IN  EXPOSURE  ASSESSMENTS

7.1  CURRENT AGENCY PRACTICE
     As stated in the introduction of this  report  the current  Agency  policy  is
to use a value of 100 percent to represent  the dermal absorption  of a penetrant
unless a lower value can be supported by scientific  studies. For  most compounds
this is a large overestimation of the actual dermal  absorption rate or percent
fraction of penetration.  In Table 4 there  are three compounds, hippuric  acid,
nicotinic acid, and thiourea, that have a total  dermal  absorption of  less than
1 percent of the applied dose after 5 days; the predicted  absorbed dose of
these three compounds would be two orders of magnitude  high  if their  dermal
absorption was assumed to be 100 percent.  On  the  other hand,  compounds like
carbaryl, caffeine, dinitrochlorobenzene, and  DMSO are  much  more  thorough
penetrants and their absorption could be fairly well approximated by  using 100
percent.  From Table 5 it appears that the  majority  of  the pesticides are
close to 10 percent for their penetration value.
                                            «
     There are many other parameters that go into  calculating  dermal  exposure and
intake besides the dermal absorption factor.   In addition, one needs  to know
the area of exposed skin, the concentration of the penetrant,  the duration of
the exposure per each event, and the frequency of  events expressed in number of
exposures per some unit time (usually per year).  The Exposure Evaluation
 Division of OTS in conjuncion with Versar, Inc.,  developed  a  general  scheme
for calculating the dermal intake of humans (Freed,  et  al. 1983). One scenario
is that of a thin film of penetrant on the  skin.   For  this  finite mass situa-
tion the exposure is calculated by multiplying (concentration) x  (skin area) x
                                      -57-

-------
(frequency) times the thickness of the film layer.   The dermal  intake  is then
calculated by multiplying the exposure times the dermal absorption  factor
expressed as a percent or as %/hr times the duration.   Another  scenario is that
of an excess amount of penetrant on the skin.  In this  case,  the thickness of
the penetrant layer is not calculated; steady-state  kinetics  are assumed and
the dermal intake is calculated by multiplying (skin area)  x  (concentration) x
(frequency) x (duration) x. (permeability constant, kp).
     Freed et al. (1983) give a sample calculation in their document showing
how the parameters above are derived for the case of ambient  human  exposure to
penetrants in water from swimming.  Information on the  frequency and duration
of outdoor swimming was found in a report by the Bureau of  Outdoor  Recreation
which was based on a survey of 11,000 people.  For this group,  34 percent  swam
in rivers, lakes, and oceans, the average frequency  of  swimming was 7  days per
year, and the average duration was 2.6 hours per day yielding a periodicity of
18.2 hours/year.  The availability for dermal exposure  in this  example was
assumed to equal the total amount of human skin surface area  which  is  17,000
cm2 for the average adult and 7,700 cm2 for the average child (1-10 yrs).  The
ambient aqueous concentration of a penetrant can be  determined  by using
monitoring data and/or by modeling.  Multiplying (skin  area)  x  (cone.) x
(frequency) x (duration) times the absorption rate for  the  specific penetrant
will give the dermal intake.
     In another study, Brown et al. (1984) looked at the role of skin  absorption
as a route of exposure for volatile organic compounds  (VOCs)  in drinking water.
They estimated absorption levels for an adult taking a  15-minute bath  and
                                      -58-

-------
drinking two liters of water, an infant  bathed  15  minutes  and  fed one  liter of
water, and a 48-pound child swimming one hour and  drinking a  liter  of  water.
Concentrations of each VOC were varied from 0.005  mg/1  of  water to  0.5 mg/1.
The authors found that for the swimmer,  a minimum  of  83 percent and possibly  91
percent of the chemicals entering the body came through the skin.   An  interesting
result is that the highest doses resulted from water  with  the  smallest concentration
of the penetrant.  Scheuplein and Blank  (1973)  have also shown the  permeation
rates are actually increased with dilute solutions as compared to pure liquids.
They illustrate this with the example of hexanol:  liquid  hexanol (8.2 M)  is
approximately 150 times more concentrated than  saturated aqueous hexanol  (0.055
M), yet the permeation rate of aqueous hexanol, far from being 150  times  less
than the pure liquid, is almost twice as great. This apparently anomalous
behavior is attributed to the compaction and dehydration of the stratum corneum
when in contact with pure liquids making it less porous, as well as to the
distribution factors of changing gradients and  partition coefficients  understood
in Pick's Law.
7.2  EXAMPLES OF DERMAL INTAKE CALCULATIONS
7.2.1  Dermal Exposure from Swimming
     There are several specific penetrant examples of how  to calculate dermal
absorption of a penetrant in swimming water.  Scow et al.  (1979) prepared a
document for the Office of Water Planning and Standards in which the exposure
levels for heptachlor and chlordane were calculated.   The  water flux (J)  through
the skin is taken to be between 0.2 to 0.5 mg/cm2-hr  while the flux of the
solute (penetrant) is estimated by multiplying  the water flux  by the weight
                                      -59-

-------
fraction of penetrant in the water.   At a  concentration  of  1  ppb,  the  flux  of
penetrant would be 0.2 x 10'9 to 0.5 x 10'9  mg/cm2-hr.   Using a  representative
body surface area of 1.8 x 10* cm2,  if one swam for 4 hours in any given  day in
water containing 1 ppb chlordane, then from  0.013  to 0.036  ug chlordane might
be adsorbed.
     Beech (1980) calculated the amount of chloroform absorbed over a  3 hour
period by a 6 year-old boy swimming  in water containing  500 ug chloroform per
                                /
liter.  The average 6 year-old is assumed  to weigh 21.9  kg  and have a  surface
area of 0.88 x 10* cm2.  The permeability  constant (kp)  for chloroform in
aqueous solution was assigned the value of 125  x 10~3 cm/hour.  The flux  through
the skin per hour is:  125 x 10~^ cm/hour  x  500 ug/liter x  1  liter/1000 cnr* x
1000 mg/1 ug = 62.5 x 10'6 mg/cm2 hour.
The total dermal intake is then:  3  hr x 62.5 x 10'6 mg/cm2 hr x 0.88  x 104 cm2
= 1.65 mg chloroform.
7.2.2  Pesticide Application
     Another area of potentially significant dermal  exposure  is  that of pesticide
application and worker reentry to areas where crops have been sprayed. Maddy
et al.  (1983) developed exposure monitoring techniques  designed to investigate
the factors influencing pesticide exposure to workers during  the application
process.  The authors conducted dermal exposure monitoring  of workers  involved
in the application of parathion, meinphos, nitrofen, DEF/Folex,  and chloroben-
zilate.  Exposures of mixers/loaders, ground applicators, mixer/loader/ground
applicators (workers performing all  three  activities during a single application),
aerial applicators, and flaggers were determined in a total  of 102 individual
                                      -60-

-------
exposure situations.  California regulations  require  use  of  clean  outer  clothing
to reduce the potential  dermal  exposure of workers  to pesticides by  decreasing
the area of bare skin available for contact with  the  chemicals.  Rubber  or
some other type of waterproof gloves were worn by all  workers  except the aerial
applicators and flaggers.
     Hand exposure was determined by rinsing  the  hands in a  predetermined
solvent containing either water, soap and water,  ethanol, or a combination  of
the three.  Hands were rinsed prior to and upon completion of  the  applications.
Exposures to the hands,  face, and neck were estimated by  placing small patches
on the upper collar of the coveralls in the front and back,  or, in some  cases,
placing patches directly on the face.  Values were  extrapolated to the entire
surface areas of these body parts.  Potential  exposure to skin protected by
coveralls was also measured with patches.
     Table 18 summarizes the average percentage of  total  dermal exposure found
on various regions by individual chemical and job activity.  The results show
that estimated exposure  to protected body areas represented  only 23.3 percent
of the total dermal exposure.  Hand exposure  exceeded exposure to  all other
areas; workers who wore  waterproof gloves still experienced  hand exposure
representing 40.9 percent of their total dermal exposure. This result indicates
why the hands are not considered to be a protected  area even though  waterproof
gloves were usually worn.  The author's possible  explanations  for  the relative
ineffectiveness of the gloves include (1) contamination of the inside material
of the gloves, (2) removal of gloves during mechanical adjustments to the
application equipment, and (3)  the handling of the  outside of  contaminated
                                      -61-

-------
gloves while putting them on or taking them off.
     Maddy et al. (1983) conclude that more attention should be given  to the
monitoring of the hands, head, face, and neck,  and less  attention  to monitoring
protected areas.  They also suggest that since  hand exposure was responsible
for more than 42 percent of the total  dermal  exposure, estimates of total
dermal exposure could be derived by multiplying the hand exposure  by a factor
of 2.5.
7.2.3  Treated Field Reentry
     A second example of dermal intake resulting  from pesticide application  is
the worker reentry study of Popendorf et al.  (1979).  Five peach orchards were
harvested each for three days at decreasing post-application intervals.   Both
aerosol and dermal exposure estimates were made for the  organophosphate pesticide
Zolone plus its metabolite Zoloxon.  Low levels of Guthion (also an organophosphate)
were present in some of the orchards from prior applications.  Aerosol  samples
were collected at 1.7 1pm for one to two hours  during the workday  from near  the
breathing zone of each worker.  Respiratory doses were calculated  from the
airborne concentrations and estimated respiratory volumes of 24 m^ during each
three day work sequence.  Dermal doses to pickers were estimated using 4x4
inch gauze patches taped to the skin.  A knit nylon glove backed by a  pad was
used to monitor hand exposure.  At the end of each week, the patches from each
location were pooled for extraction and analysis.  The resulting value was
adjusted for exposure time and patch area at each location and then multiplied
by its proportionate body surface area; these were summed to estimate  the whole
body dermal dose of each compound for each week (see Table 19).  Approximately
                                      -62-

-------
                            TABLE 18A.  RELATIVE CONTRIBUTIONS TO TOTAL DERMAL EXPOSURE OF BODY AREAS
                          TO PESTICIDES AS STUDIED BY THE CALIFORNIA DEPARTMENT OF  FOOD AND AGRICULTURE
   Chemical

Parathi on
Mevinphos
TDK
DEF/Folex
Chi orobenz. late

Total
            Average Percentage
Number of     of Total  Dermal
Exposures     Exposure  Found
Monitored       on Hands
   3
  22
  24
  32
  21

 102
23.4
48.0
49.2
46.8
27.1

42.9
Average Percentage
 of Total Dermal
Exposure Found on
Head. Face and Neck

      67.4
      34.6
      22.7
      23.7
      59.4

      33.8
Average Percentage
 of Total  Dermal
Exposure Found on
  Hands and Head,
  Face and Neck

      90.8
      82.6
      71.9
      70.5
      86.5

      76.*7
Average Percentage
 of Total Dermal
Exposure Found on
  Protected Area

       9.2
      17.4
      28.1
      29.5
      13.5

      23.3
                    TABLE 188.  RELATIVE CONTRIBUTIONS TO TOTAL DERMAL  EXPOSURE  OF  BODY AREAS BY JOB ACTIVITY




Job Activity
Mixer/Loader.
Ground App'. icator
Mixer/Loader
Aerial Applicator
Ground Appi Icator
Fl agger
Total


Number of
Exposures
Monitored

4
36
18
25
19
102

Average Percentage
of Total Dermal
Exposure Found
on Hands

18.1
50.7
54.6
30.4
38.7
42.9

Average Percentage
of Total Dermal
Exposure Found on
Head, Face and Neck

57.5
22.0
27.4
47.9
38.6
33.8
Average Percentage
of Total Dermal
Exposure Found on
Hands and Head,
Face and Neck

75.6
72.7
82.0
78.3
77.3
76.7

Average Percentage
of Total Dermal
Exposure Found on
Protected Areas

24.4
27.3
18.0
21.7
22.7
23.3
From Maddy et al. 1983

-------
98-99% of the total  doses  thus calculated  were  attributed to the  dermal  rather
than respiratory route. These calculations  do  not  account  for  possible  variations
in the rate of chemical absorption  at  different skin  locations.
7.2.4  Hypothetical  Dermal Intake of PCBs
     Versar, Inc. (1983a)  in conjunction with the Exposure  Evaluation Division
of OTS has estimated maximum probable  dermal  intake to  PCBs for several
scenarios.  The first example deals with dermal exposures in the  occupational
environment.  Annual dermal  intake  of  PCBs resulting  from use of  a  specific
chemical  or product  (assuming that  no  protective clothing or equipment is worn)
can be estimated from
     Amount PCBs     PCBs  available         Frequency of
      absorbed    =   for absorption    x      exposure      x    Absorption
       (mg/yr)         (mg/event)           (events/yr)             (%)
where, for liquids,  PCBs available  for absorption =TxLxCxS
     -  T (liquid film thickness) is assumed  to be  0.0018 cm.   This  is the
        average of the measured film thicknesses of five solutions  on the skin
        after immersion of hands into  the  solution  followed by  a  partial wipe
        with a rag:   mineral oil, cooking  oil,  bath oil, 50% bath oil/50% water,
        and water (Versar, 1983b)
     -  L (density of liquid) is assumed to  be  1.6  x  ID-* mg/cm^.
     -  C is the PCB concentration  in  the  liquid (kg/kg)
     -  S (skin area exposed) is assumed to  be  the  entire surface area of both
        hands which  is taken to be  870 cm2.
and, for dusts, PCBs available for  absorption = M x C x S
     -  M is the maximum mass of a  dust that  can adhere to  one  cm2  of skin
        which is 2.77 mg/cm2 (Versar,  1982)
                                      -64-

-------
                           TABLE  19.  WORKER REENTRY EXPOSURE
                INTEGRATED DERMAL DOSE TO EACH LOCATION, ALL VALUES IN mg
Zolone
                                                   Week
Hands
Forearms
Upper arms
Head
Neck
Shoulders
Chest
Back
Hips
Th1 ghs
Calves
Feet
Total without hands
Overall

Hands
Forearms
Upper arms
Head
Neck
Shoulders
Chest
Back
H1ps
Thighs
Cal ves
Feet
Total without hands
Overall
95.0
9.8
4.3
7.3
1.6
0.4
1.0
0.6
0.2
0.7
1.4
0.2
27.3
122.3
Gut Mont
20.8
2.82
0.57
1.38
0.30
0.06
0.24
0.06
0.04
0.14
< 0.02

5.62
26.4
116.1
17.6
14.5
11.6
2.4
1.7
1.8
1.7
0.2
1.0
0.3
0.1
52.9
169.0
Zoloxon
2.17
0.28
0.17
0.17
0.03
0.04
< 0.01
< 0.01
< 0.01
< 0.01
< 0.01
< 0.01
0.69
2.86
84.2
12.2
13.4
23.0
2.4
1.4
1.1
2.0
0.6
2.3
0.5
0.1
58.9
143.1
Zoloxon
1.82
0.29
0.35
0.68
0.05
0.04
0.03
0.05
0.02
0.06


1.56
3.38
158.0
12.3
13.1
24.1
2.6
. 0.7
2.5
3.8
1.3
5.1
7.9
0.9
74.3
232.1
Zoloxon
1.39
0.08
0.11
0.22
0.04
0.01
0.03
0.10
0.02
0.07
0.01
0.01
0.74
2.13
tA total dose of 2.6 mg Zoloxon  (2.5  to the  hands) in week 1 Is not listed.
From Popendorf et al. 1979-
                                          -65-

-------
     -  C is the PCB concentration  in  dust  (kg/kg)
     -  Sis the skin area exposed  per event  (cm^/event)
     A sample calculation for the exposure  scenario  of  loading/unloading  a
liquid containing 50 mg/kg PCB with a  frequency  of 96 times/yr  (with absorption
assumed to be 100%)  is:
     Amount PCBs  =   TxLxCxSx96 events/yr. x 100%
   absorbed (mg/yr)
                  =   0.0018cm x 1.6 x  103 mg/cm3 x 50 mg/kg  x 870cmz x  96
                     events/yr
                  =   12  mg/yr
     The second example  is the calculation  of the maximum  probable  individual
dermal intake to PCBs resulting from routine  use of  general  household cleaners
containing incidental  PCBs.  The dermal intake calculations  apply to typical  use
of a solid household detergent designed to  be mixed  with water.  The following
assumptions were made:
     -  96 gm of detergent is used  per cleaning  job  and detergent is mixed  with
        1 gallon (3.785  1) of water.
     -  Detergent contains 25 percent  by weight  of the  PCB - contaminated
        chemical.
     -  A 0.0024cm thick film of the cleaning solution  remains  on the hand  and
        part of the  forearm, covering  a skin  surface area  of 500 cm2, after
        each immersion.   This is the measured film thickness for a  solution of
        50% water/50% bath oil retained on  the skin  after  immersion of  hands
        into the solution followed  by  a partial  wipe with  a  rag  (Versar,  1983b).
        Therefore, 0.0024 cm x 500  cm2 =1.2  cm3 (ml) of solution remain  on the
        skin.
                                      -66-

-------
     -  Detergent is used once a week for a total  annual  frequency of 52  times
        a year.
     -  One hand is immersed 20 times in the bucket during weekly cleaning
        sessions and no gloves are worn.
     -  All PCBs in the solution remaining on the  skin after each immersion are
        available for dermal absorption.
The PCBs available for absorption (mg/event) =AxBxCxD
     A = PCB in constituents (mg/kg)
     B = Aqueous dilution of detergent; quantity detergent per volume of  water
         (kg/ml)
     C = Weight fraction of PCB - contaminated constituent
     D = Volume of solution on skin per immersion  (mg/event)
Frequency = 20 immersions per week x 52 weeks per  year = 1,040 events/yr
Absorption = 100 percent
     A sample calculation for PCB concentration of 50 mg/kg is :
Amount PCBs =AxBxCxDx Frequency x Absorption
 absorbed
            = 50 mg x    .096 kg     x .25 x 1.2   ml_  x 1,040 events
               kg     3.785 x 103 ml             event           yr
            = .4 mg/yr
                                      -67-

-------
                CHAPTER 8 - ANALYSIS/FUTURE RESEARCH NEEDS

     The most direct method of obtaining percutaneous absorption data for a
penetrant is to conduct a human in vivo study.  We have seen, however, that
there are many problems with this direct approach such as toxicity concerns,
having to monitor urine output for 21 days for slow penetrants, and the 5-fold
variation possible between subjects.  Kligman (1983), in a recent review of
percutaneous absorption, states in comparing in vivo to in vitro results that
in vitro data are more credible on technical grounds alone as the methods are
more precise, involve less experimental error, and the variables are under much
greater control.  From the limited number of in vitro results presented in this
review, it appears that the in vitro studies are more reproducible than j_n_
vivo studies.  Because of the problems associated with in vivo studies, it is
'suggested that research be focused on compiling absorption data with in vitro
studies using 1 type of skin which then can be compared to the permeability of
skin from all other parts of the body (i.e., adominal skin could be tested and
ratios such as found in Table 2 could be used to determine absorption at other
sites like the forearm or palm).  The in vitro results could then represent
actual penetration in vivo; however, it may be prudent to include a "safety
factor" of 5 (or other number derived from more extensive studies) to account
for the variation between subjects as noted in the in vivo studies.
     The previous chapters have shown the importance of the stratum corneum/water
partition coefficient, Km.  Unfortunately, there is only a limited data set
of measured Km values, thus its utility as a predictor of dermal absorption
is largely unknown.  Furthermore, the accuracy of measured Km values is questioned
                                      -68-

-------
by Scheuplein et al. (1969)  who state that  partition  coefficients  like  Kamyi
caproate (^ac) or ^hexane can &e obtained more accurately  than  Km.
     It is suggested that a  program to measure Km for a  large number  of
chemicals and covering a variety of functional groups be established.   In  this
manner we could determine how well  in vitro Km measurements  represent the
actual in vivo partition between the penetrant and skin.  Also, we would like
to know if Km can be calculated from other  partition  expressions like Kac
°r Kolive oil*  For example, a regression analysis was done  for the
                                    b
linear alcohol series giving Km = aK0iive 01-]  with r  .= .97,  a = 3.748,
and b = .448 showing a good  correlation for this  series.  However,  a  similar
regression analysis for the  steroids (the only other  chemicals  with measured
Km values) had a poorer fit  with r = .78, a = 8.742,  and b = .4288, possibly
due to inaccuracies in the measurement of Km.
     The second parameter of importance needed to calculate  permeability
is the diffusivity (D).  We  have see that for the linear primary alcohols
D is relatively constant (D  = 10"^ cm^ sec"*), while  for the series of
steriods, D is not constant, varying by approximately 200-fold.  This
variation can be explained in part by changes  in  molecular volume and polar
functional groups.  As with  Km, it is suggested that  D values be tabulated
for a number of chemicals covering a variety of functional groups  so  that
the dynamic range of D can be determined.  It  may be  possible to approximate
D for an unknown chemical by comparison to  chemicals  with  measured D  values.
     The compilation of Km and D values for a number  of  representative
chemicals will facilitate the estimation of kp for unknown chemicals.   In
                                      -69-

-------
 order  to  promote consistency  in the compilation of these values, it would
 be prudent  to  decide what  tissue  (man or animal) to use in diffusion
 cells  and partition measurements  of Km plus from what body site.  If
.animals are to be  used  either for in vivo or in vitro studies, it would
 be most helpful for comparative purposes to decide what animal species is
 closest to  man and use  only the species chose.  The rhesus monkey and
 miniature swine are closest in absorption properties to man, but the rat
 may be the  species of choice  due  to economy and ease of handling.  Also,
 it is  suggested that clearcut procedures be developed to measure Km
 and D  values for penetrants of low water solubility.  In vitro diffusion
 cell measurements  are currently restricted to penetrants that are at
 least  partially water solubile, although some effort is underway to use
 sol vent/water  mixtures  for the receptor bath.
     Values for Km and  D are  required to calculate the permeability when
 there  is  an excess of penetrant and steady-state kinetics are followed.
 However,  there are other scenarios, such as when a finite amount of
 penetrant (i.e., thickness of penetrant layer is smaller than the thickness
 of the stratum corneum) is applied and the penetrant is fast; Fickian
 diffusion kinetics may  not be followed for this case.  It would be useful
 to establish expressions for  kinetic flux for all possible scenarios such
 as:  finite amount of a fast  penetrant, finite amount of a slow penetrant,
 excess of a penetrant where the diffusion through the stratum corneum is
 not rate  limiting, buildup of penetrant in blood so that C0 7* 0, etc.  In
 this manner, once  the appropriate scenario is determined from the exposure
                                    -70-

-------
situation, the proper kinetic expression  can  be  used  to  determine  the  flux.
     The use of partition systems to predict  dermal  absorption  has been
discussed in the previous chapters.   Partition coefficients  themselves
have been predicted from structure-activity relationships,  number  and
kind of functional  groups, and by other techniques  (Sato and Nakajima  1979,
Hansch et al. 1972  and 1973, Fujita  et al.  1964, Leo  and Hansch 1971a  and
1971b, and Katz and Shaikh 1965). Thus,  it may  be  possible  to  use the
methods developed to predict partition coefficients to calculate dermal
absorption once the relationship between  a  partition  coefficient and
permeability is established.
     We have seen in the previous chapters  that  there are two approaches
used to represent the amount of penetrant diffusing through  the stratum
corneum.  The first and most common  method  is to use  the total  percent
adsorption calculated or estimated (such  as 100% absorption); this method
does not take into  account the contact time or the  particular kinetics
that may apply depending on the speed of  the  penetrant,  its  thickness  on
the skin, or if it  is bound to the vehicle  in some  manner.   The second
method is to use an absorption rate  which then has  to be coupled with  a
contact time.  It would be valuable  to review the advantages/disadvantages
of these approaches, particularly in terms  of their relative uncertainty.
     Since the conservative approach when no  data is  available  for a pene-
trant is to use 100 percent absorption, it  may be possible  to group
penetrants into a numerical system such as  100%  - 10% -  1%  - .1% absorption
depending on physical parameters such as  Km and  D,  on partition coefficients,
                                   -71-

-------
or in vitro tests.   This  grouping of  penetrants into a numerical system
based on physical  parameters  is  similar  in  intent to the grouping of
steroids by their  Kac value as is done by Scheuplein et al. (1969)
(see Chapter 6).  The use of  a numerical system is an "order of magnitude"
approach which may be justified  when  the uncertainty of all the factors
leading to dermal  intake  are  taken  into  consideration.
                                   -72-

-------
                               9.   REFERENCES


Anderson, S.L., and Cassidy,  J.M.   1973.   Variations in physical dimensions and
     chemical  composition of  human  stratum corneum.  J. Invest. Dermatol.  61:
     30-32.

Bartek, M.J.,  LaBudde, J.A.,  and Mai bach,  H.I.   1972.  Skin permeability
     in vivo;  comparison in rat, rabbit, pig, and man.  J. Invest. Dermatol.
     58:114-123.

Bartek, M.J.,  and LaBudde, J.A. 1975.  Percutaneous Absorption In Vitro.  In:
     H.I. Maibach (ed), Animal  Models in Dermatology.  ChurchhilT~Livingstone,
     New York, p. 103.

Beech, J.A.  1980.  Estimated worst case trihalomethane body burden of a child
     using a swimming pool.  Medical Hypotheses.  6:303-307.

Bronaugh, R.L., Stewart, R.F.,  Congdon, E.R., and Giles, A.L.  1982.  Methods
     for in vitro percutaneous  absorption  studies.  I. Comparison with in vivo
     results"Tox. Applied Pharmacol.  62:474-480.

Brown, H.S., Bishop, D.R., and  Rowan, C.A.  1984.  The role of skin absorption
     as a route of exposure for volatile organic compounds in drinking water.
     Am. J. Public Health 74:479-484.

Dugard, P.H.  1983.  Skin permeability theory in relation to measurements of
     percutaneous absorption  in toxicology.  In:  F.N. Marzulli and H.I. Maibach
     (ed.), Dermatotoxicology.  Hemisphere Publishing Corporation, Washington,
     p. 102.

Eli as, P.M., Cooper, E.R., Korc, A., and Brown, B.E.  1981.  Percutaneous
     transport in relation to stratum corneum structure and lipid composition.
     J. Invest. Dermatol. 76:297-301.

Feldmann, R.J. and Maibach, H.I.   1969.  Percutaneous penetration of steroids
     in man.  J. Invest. Dermatol., 52:89-94.

Feldmann, R.J., and Maibach,  H.I.   1970.   Absorption of some organic compounds
     through the skin in man.  J.  Invest.  Dermatol., 54:399-404.

Feldmann, R.J., and Maibach,  H.I.   1974.   Percutaneous penetration of some
     pesticides and herbicides  in  man.  Tox. Applied Pharmacol.  28:126-132.

Flynn, 6.L.  1971.  Structural  approach to partitioning:  Estimation of steroid
     partition coefficients based  upon molecular constitution.  J. Pharma. Sci.
     60:345-353.
                                      -73-

-------
Franz, T.J.  1975.  Percutaneous  absorption.  On the relevance of in vitro data.
     J. Iniest. Dermatol.   64:190-195.

Franz, T.J.  1978.  The finite  dose technique as a valid in vitro model for the
     study of percutaneous absorption  in  man.  Curr. Probl. Dermatol.  7:58-68.

Freed, J.R., Chambers,  T., Christie, W.N.,  and Carpenter,  C.E.  1983.  Methods
     for assessing exposure to  chemical substances.  EPA 560/5-83-015.  U.S.
     Environmental Protection Agency,  Washington, DC, 1983.  407 pp.

Fujita, T., Iwasa, J.,  and Hansch, C.   1964.  A new substituent constant, n,
     derived from partition coefficients.   J. Amer. Chem.  Soc.  86:5175-5180.

Guy, R.H. and Maibach,  H.I.  1984.  Correction factors for determining body
     exposure from forearm percutaneous absorption data.   J. Appl. Toxicol.
     4:26-28.

Hansch, C., Leo, A., and Nikaitani, D.  1972.  On the additive-constitutive
     character of partition coefficients.   J. Org. Chem.   37:3090-3092.

Hansch, C., Leo, A., linger, S.H., Kim,  K.H., Nikaitani, D., and Lien, E.J.
     1973.  Aromatic substituent  constants  for structure-activity correlations.
     J. Med.  Chem.  16:1207-1216.

Higuchi, T.  1960.  Physical  chemical  analysis of percutaneous absorption
     process from creams and ointments.   J. Soc. Cosmet. Chem.  11:85-97

Higuchi, W.I.  1962.  Analysis  of data  on the medicament release from ointments.
     J. Pharm. Sci.  51:802-804.

Katz, M. and Shaikh, Z.I.   1965.  Percutaneous corticosteroid absorption
     correlated to partition coefficient.   J. Pharm. Sci.  54:591-594.

Kligman, A.M.  1983.  A biological brief  on percutaneous absorption.  Drug
     Devel. Indust. Phar.   9:521-560.

Krueger, G.6. and Shelby,  J. 1981.  Biology of human skin transplanted to the
     nude mouse:  I. Response to  agents which modify epidermal proliferation.
     J. Invest. Dermatol.   76:506-510.

Lien, E.J. and Tong, G.L.   1973.  Physiocochemical properties and percutaneous
     absorption of drugs.   J. Soc. Cosmet.  Chem.  24:371-384.

Leo, A., Hansch, C.  1971.  Linear free-energy relationships between partitioning
      solvent systems.   J. Org. Chem.   36:1539-1544.

Leo, A., Hansch, C., and El kins,  D.  1971.  Partition coefficients and their
     uses.  Chem. Rev.   71:525-616.
                                      -74-

-------
Maddy, K.T., Wang, R.G.,  and Winter,  C.K.   1983.  Dermal exposure monitoring of
     mixers, loaders, and applicators of pesticides  in California.  HS-1069,
     Worker Health and Safety Unit, State  of  California, p.1-7.

Maibach, H.I., Feldmann,  R.J.,  T.H.  Milby, W.F.  Serat.  1971.  Regional variation
     in percutaneous penetration in man.  Arch. Environ. Health  23:208-211.

Michaels, A.S. Chandrasekaran, S.K.,  and Shaw, J.E.   1975.  Drug permeation
     through human skin:   Theory and  in  vitro experimental measurement.  AIChE 21;
     985-996.

Moore, W.J.  Physical Chemistry.  1962.   Prentice-Hall, Inc., Englewood Cliffs,
     NJ, 844 pp.

Ostrenga, J., Steinmetz,  C., and Poulsen,  B.  1971.   Significance of vehicle
     composition I:  Relationship between  topical vehicle composition, skin
     penetrability and clinical  efficacy.   J. Pharm.  Sci.  60: 1175-1183.

Pannatier, A., Jenner, B. Testa, B.,  and Etter, J.C.  1978.  The skin as a drug-
     metabolising organ.   Drug Metab. Rev. 8: 319-343.

Popendorf, W.J., Spear, R.C., Leffingwell, J.T.,  Yager, J., and Kahn, E.  1979.
     Harvester exposure to Zolone (phosalone) residues in peach orchards.
     J. Occup.  Med.  21:189-194.

Rongone, E.L.  Skin structure, function, and  biochemistry.  In:  F.N.
     Marzulli and H.I. Maibach (ed.), Dermatotoxicology.  Hemisphere Publishing
     Corporation, Washington, 1983.   p.2.

Sato, A. and Nakajima, T.  1979.  A structure-activity relationship of some
     chlorinated hydrocarbons.  Arch. Environ. Health 34:69-75.

Schaefer, H., and Schalla, W.  1980.   Kinetics of percutaneous absorption of
     steroids.  In:  P. Mauvais-Jarvis,  D.F.H. Vickers, and J. Wepierre (ed.),
     Percutaneous absorption of steroids.   Adademic  Press, New York. p.54.

Scheuplein, R.J.  1965.  Mechanism of percutaneous absorption.  I: Routes of
     penetration and the  influence of solubility.  J. Invest. Dermatol.  45:334-346,

Scheuplein, R.J.  1967.  Mechanism of percutaneous absorption.  II. Transient
     diffusion and the relative importance of various routes of skin penetration.
     J. Invest. Dermatol.  48:79-88.

Scheuplein, R.J., Blank,  I.H., Brauner,  G.J., and MacFarlane, D.J.  1969.
     Percutaneous absorption of steroids.   J. Invest. Dermatol. 52:63-70.

Scheuplein, R.J. and Blank, I.H.  1971.   Permeability of the skin.  Physiol.
     Reviews. 51:702-747.                                          	
                                      -75-

-------
Scheuplein, R.J. and Blank,  I.H.   1973.   Mechanism of percutaneous absorption.
     III.  Penetration of nonelectrolytes (alcohols) from  aqueous solutions and
     from pure liquids.  J.  Invest.  Dermatol.   60:286-296.

Scheuplein, R.J., and Ross,  L.W.   1974.   Mechanism of percutaneous absorption.
     V.  Percutaneous absorption  of  solvent  deposited solids.  J. Invest. Dermatol.,
     62:353-360.                                                 !

Scheuplein, R.J.  1980.  Percutaneous  absorption:  Theoretical aspects.  In:  P.
     Mauvais-Jarvis, C.F.H.  Vickers, and  J.  Wepierre (ed.), Percutaneous "absorption
     of steroids.  Academic  Press, New York, p.9.

Scow, K., Wechsler, A.E., Stevens, J., Wood, M., and Callahan, M.  1979.
     Identification and evaluation of  waterborne routes of exposure from other than
     food and drinking water.   EPA-440/4-79-016, U.S. Environmental Protection
     Agency, Washington, DC. 58 pp.

Shah, P.V., Monroe, R.J., and  Guthrie, F.E.  1981.  Comparative  rates of dermal
  '   penetration of insecticides  in  mice. Tox. Applied Pharmacol. 59:414-423.

Stoughton, R.B., Clendenning,  W.E.,  and Kruse,  D.  1960.   Percutaneous absorption
     of nicotinic acid and derivatives.   J.  Invest. Dermatol.  35:337-341.

Versar.  1982.  Priority review level  I exposure assessment for  MOCA.  Final
     report.  U.S. Environmental  Protection  Agency, Washington,  DC, Office of
     Toxic Substances, Contract No.  68-01-6271.

Versar.  1983a.  Exposure assessment for  incidentally produced PCBs.  Draft final
     report.  U.S. Environmental  Protection  Agency, Washington,  DC, Office of
     Toxic Substances, Contract No.  68-01-6271.

Versar.  1983b.  Exposure assessment for  retention of chemical liquids on hands.
     Preliminary draft report.  U.S. Environmental Protection Agency, Washington,
     DC, Office of Toxic Substances.  Contract  No. 68-01-6271.

Wester, R.C., and Maibach, H.I.  1975. Percutaneous absorption  in the rhesus
     monkey compared to man.  Toxiocol. Appl. Pharmacol.   32:394-398.

Wester, R.C., Noonan, P.K.,  and Maibach,  H.I.   1977.  Frequency  of application
     on percutaneous absorption of hydrocortisone.  Arch.  Dermatol  113:620-622.

Wester, R.C., and Noonan, P.K.  1980a. Relevance of animal models for percutaneous
     absorption.  Intern. J. Pharm.  7:99-110.
                                      -76-

-------
Wester, R.C., Noonan, P.K., and Mai bach,  H.I.   1980b.   Percutaneous  absorption
     of hydrocortisone increases with long-term administration.   Arch. Dermatol.
     116:186-188.

Wester, R.C., and Mai bach, H.I.  1983.  Percutaneous pharmacokinetics:   10  Steps
     to percutaneous absorption.  Drug Metab.  Rev.  14:169-205.
                                      -77-

-------