70 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS will be obtained only if the penetrating substance does not combine with any component of the skin, and if the skin circulation is fully effective in clearing away the material which has entered the skin. The practical in rai**o methods that can be employed are somewhat limited. If the test substance is pharmacologically active, the body responses can be studied quantitatively. This is not easy and usually not very precise. Furthermore, the degree of a physiological response to a drug is rarely, if ever, directly proportional to the dose of the drug. Some sort of calibration of the body response is therefore necessary. If, instead of directly attempting to measure the quantity absorbed into the body, the disappearance of the applied test substance is studied, then it is necessary to employ an assay method which can detect a very small change in a relatively large quantity. If chemical analysis is used, there is also the difficult problem of removing the residue of test substance from the skin surface for assay purposes. This procedure must be repeated at various times after applying the substance and can only be carried out on different regions of the skin. Many of these difficulties disappear if radioisotopes are employed as trace elements, but some problems still remain. Estimates of absorption of the labeled test substance into the body must rely on measurements of the rate of excretion of the trace element or changes in the circulating blood concentration of the tracer. But these measurements alone are meaning- less and must be supported by studies of distribution, excretion, and so on. If the test substance is injected intravenously at a constant rate, then the blood concentration of the substance may be usefully correlated with the injection rate, or with the integrated injection rate--that is, the total dose absorbed. This approach has been successfully employed by Griese- met, Blank and Gould (2) in their'studies of the percutaneous absorption of Satin. With this particular substance no trace element was necessary because the Satin irreversibly reduced the level of the circulating blood cholinesterase. It is also feasible with small animals to determine the total dose of radioactive tracer absorbed by destroying and dissolving the experimental animal, and estimating the radioactivity of the resulting solution. An obviously necessary first procedure is the removal of the skin to which the test substance had been applied. This is a static determination, but a sufficiently large series of such experiments carried out at various times after applying the substance to the skin would elucidate the time-course of the skin penetration. It is clear that some of the methods mentioned could not be used for studying the penetration of human skin. However, it would seem possible to administer an innocuous radio-labeled test substance intravenously to humans and obtain a correlation between the blood concentration of radio-

METHODS FOR MEASURING PERCUTANEOUS ABSORPTION 71 activity and the rate of injection. This relation might then be used to estimate the rate of percutaneous absorption from the blood radioactivity. But there are serious limitations, in particular to the degree of radioactivity which could be used. Suppose 1 per cent of a dose applied to the skin penetrates and becomes evenly distributed in 50 litres volume of tissue. Suppose also that we take a 10 mi. sample of blood for assay and that the minimum rate of counting is one count per second for reasonable precision. Then allowing 10 per cent efficiency for the Geiger counter, we can calculate that the activity of the test substance applied to the skin would have to be more than 100 microcuries. This is a high level of radiation activity which, with a •-emitter of high energy such as P32, would probably produce a severe skin reaction. A safer way of studying the percutaneous absorption of a radioactive compound is to follow the disappearance of the compound from the skin starface. This can be carried out with a very small quantity of the test substance of a total radioactivity less than 1 microcurie. It can be said, therefore, that attempts to estimate percutaneous absorp- tion in vivo have to be based on indirect measurements and, in general, difficult. and not very precise techniques. A more direct experimental approach can be made by directly measuring the rate of diffusion through isolated skin. But first we need evidence that removal of skin from the body does not alter any characteristics which may modify its permeability. Treherne (1) found that removal of the epidermis increased the permeability of rabbit skin by two to three orders, and Blank, Griesemer and Gould have shown by autoradiography (3) that shallow cuts penetrating just a little deeper than the stratum comeurn allow rapid penetration. It is likely that the main barrier to chemical diffusion through the skin is in the epidermis and may be localized near the base of the stratum corneum. We are, therefore, concerned mainly with the integrity of the epidermis in diffusion studies on isolated skin. When the skin is being •esected, the largest factor likely to interfere with its integrity is cutting off its blood supply. But the epidermis does not normally have a blood supply and any changes in it due to cutting off the skin circulation are, therefore, not likely to be instantaneous or even rapid. Treherne, in fact, found that freshly resected whole skin respired oxygen TAI•LE 1.--THE EFFECT oF STORAGE ON THE STEADY RATE OF PENETRATION OF ISOLATED SKIN BY TRI-BUTYL PHOSPHATE Freshly re- sected skin, Stored Stored Species w/min./cm? for 24 hr. for 48 hr. Rabbit 1.16 1.20 1.48 Pig 0.23 0.26 ... Man 0.16(8hr.) 0.22 ... at a rate which was practically un- changed several hours later. Our experiments have indicated that if bacteria] decomposition of resected skin is prevented by cold storage, the chemical permeability of the skin at 33øC., is almost unchanged for at least two days. Direct comparison