370 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS (2) and Manowitz and Johnston (12) used organic solvents to extract anti- bacterial substances deposited on skin and analyzed the extracts spectro- photometrically. Fahlberg et al. (9) assayed the extracted agent micro- biologically. Investigating material deposited in particles large enough to be observed directly, Parran (13) stripped the skin with sticky tape and under magnification observed the amount of particles adhering to the tape. Shemano and Nickerson (14) and Stoughton (15) used radioactive hexachlorophene and measured the quantity sorbed onto the skin by sur- face counting. Using radioactive soaps, Hopf and Burmeister (16) ex- tracted the skin and examined the extracts for the amount of soap re- tained after washing. In the work reported here, which extrapolates the method of Shemano and Nickerson (14) to humans, the accumulation and persistence of a mixture of antibacterial substances deposited from soap onto human skin were studied. It is well known that antibacterial cleansing products such as hexachlorophene soaps or detergents exert their effect in a cumulative manner, the bacterial population decreasing with frequency of use until it is more or less stabilized at a level which is characteristic for each per- son (12, 17-20). Our work indicates that the amount of antibacterial agents on the skin approaches equilibrium. The composition employed in this study consisted of a soap solution containing a 1:1 mixture of hexachlorophene (2,2'-dihydroxy-3,3',5,5',6,6'- hexachlorodiphenylmethane) and triclocarban (TCC, 3,4,4'-trichlorocar- banilide), * the former compound being C-14 labeled in the methylene group and the latter C-14 labeled in the carbonyl group. A 1:1 combina- tion of hexachlorophene and triclocarban is found in a widely used anti- bacterial-deodorant soap. METHODS The experimental work included both animal and human studies. Albino rats were used to establish a relationship between skin surface counts and absolute radioactivity present in and on the skin. It was as- stimed that this relationship also would apply to the human studies. Two male and two female albino rats weighing 250 to 300 g were em- ployed. An area on the back (approximately 6 in. a) was closely clipped and depilated* to effect complete removal of hair. Chemicals were weighed to the nearest 0.1 mg and transferred quan- titatively to a 2% soap solution the formulations were homogenized in a * Active ingredients in Dial© soap, Armour-DiM, Inc., Chicago, Ill. t Nair, Carter Products, Inc., Ncw York, N.Y.

ANTIBACTERIAL AGENTS IN HUMAN SKIN 371 tissue homogenizer equipped with a tight-fitting pestle until successive aliquots of the suspension gave counts reproducible to within 2% or less. Before use, the formulation was warmed to 30øC and agitated continu- ously with a magnetic stirrer. The composition of the test solution is given in Table I. The activity of the preparation was determined by scintillation counting of aliquots after homogenization. Table I Composition of Test Solution Ingredients g/100 ml Hexachlorophene Labeled 0.400 • 311.5 /•c Unlabeled 0. 350 TCC Labeled 0.092 • 311.5 /•c Unlabeled 0. 658 Ivory Soap a 2.0 Total gc/100 ml 623.0 Procter & Gamble Co., Cincinnati, Ohio. The formulation was applied 24 hours after depilation and after in- spection of the animals to insure absence of abrasions or irritation. Prior to application of test material, the rats were anesthetized and circles hav- ing a diameter of 3.5 cm were circumscribed on each prepared skin site with waterfast ink. A 0.1-ml sample of the test solution was spread evenly over the test site by rubbing briskly with the tip of the delivery pipet for $ min. At various intervals after application, ranging from 0 to 8 min in order to permit deposition of variable quantities of material on the skin, the test sites were thoroughly rinsed with tap water to remove excess test material. Immediately after rinsing, each animal was sacrificed and the skin at each application site was excised, spread on aluminum foil, and counted using a Nuclear Radiation Detector Model D-34' (mica window mass of 1.4 mg/cm 2 with an effective diameter of 2.778 cm) connected to a Nu- clear-Chicago Analyzer/Scaler 8725.* Triplicate skin samples, each having an area of 0.363 cm 2, were taken from each counting area using a cork borer. These samples were dried and burned using the Schoeniger combustion method. The radioactive * Nuclear Chicago, Des Plaines, Ill. 60018.