506 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS with a T-emitting isotope of Na as a tracer in order to study its absorption from four types of ointment formulations applied to the forearm of human volunteers. The ointments were anhydrous lanolin, white petroleum jelly, lard, a 'hydrophile ointment base of the cholesterol type' and a 'washable' ointment consisting of cetyl alcohol, white wax, propylene glycol, sodium lauryl sulphate and 72•o water. Observations over a 24 h period showed that absorption was best from anhydrous lanolin, next from lard, then from the hydrophile ointment and cetyl alcohol base and least of all from petroleum jelly. A better absorption of vitamin A in rats from an ointment containing lard than from one in which lard was substituted by petrolatum was reported by Sobel, Parnell, Burton, Sherman and Bradley (155). The more effective absorption of an electrolyte and an oil-soluble substance from lanolin and lard in these experiments, tended to confirm the reports of earlier workers. The effects of ointment bases in the form of w/o and o/w emulsions on percutaneous absorption have been investigated. Shelmire (20, 156), using skin irritation as a criterion for the penetration through the stratum corneum, studied the effect of incorporating croton oil (5•o), salicylic acid (2.5•o), mercury bichloride (1.5•o), resorcinol (5•o) and ephedrine hydrochloride (2•o) into petrolatum, a w/o emulsion, and o/w emulsion or Carbowax 1500 (a water-soluble ointment made up of equal quantities of polyethylene glycol 300 and 2540) and left in contact with the skin for 3 h. The oil-soluble croton oil and ephedrine produced a maximum degree of irritation when incorporated in petrolatum. The intensity was less when the w/o emulsion was used as an ointment base and least of all when the o/w emulsion was used for this pur- pose. No irritation occurred when Carbowax 1500 was used a sa base. On the other hand, the water-soluble substances salicylic acid, mercury bi- chloride or resorcinol were most effective when incorporated into the oint- ment base consisting of an o/w emulsion, other ointment bases gave a weak or no reaction. Similar results were obtained by Wahlberg (157) in his studies on the absorption of metallic test compounds from ointment bases. Experiments with other compounds did not show such a clear distinction. Thus, Barrett, Hadgraft and Sarkany (158) showed that methyl nicotinate is absorbed equally well from an o/w or a w/o cream. Studies on the percutaneous absorption of betamethazone 17-valerate and of fluocinolone acetonide (153) showed that vasoconstriction (assessed from the area of pallor), produced by the preparation containing the valerate in Carbowax 1500 was strikingly larger than that produced by

PERCUTANEOUS ABSORPTION 507 preparations in which the valerate was incorporated in aqueous cream BP, oily cream BP, and white soft paraffin BP. The reverse effect was found when fluocinolone acetonide was used in these vehicles instead of the valerate. The most marked vasoconstrictor effect was observed when fluocinolone acetonide was incorporated in aqueous cream BP and in the oily cream, it was less with white soft paraffin as an ointment base and markedly less with Carbo•vax 1500. Further work carried out by the authors (159) showed that the physical state of the steroid may have been responsible for this difference. Thus, using the microcrystalline rather than the coarse granular form improved skin absorption from white soft paraffin so that the degree of vasoconstriction produced was equal to that observed with the steroid incorporated in aqueous or oily cream BP. The improved absorption which accompanied this change in the physical state may be due to an improved contact with the skin surface since incorporating the steroid dissolved in 5•o propylene glycol, thus improving still further skin contact, increased skin penetration from these two ointments and still further from white soft paraffin. Christie and Moore-Robinson (160) also found that fluclorolone acetonide produces a better vasoconstrictor effect when dissolved in propylene glycol and then incorporated into a 'petrola- tum' ointment base. They found furthermore that the presence of 15•o or 30• cetyl alcohol further increased the vasoconstrictor effect. According to the authors the addition of cetyl alcohol has the effect of increasing the concentration of the steroid in the propylene glycol and since it is this liquid phase that is likely to have the most intimate contact with the skin, enhanced penetration of the steroid is to be expected. Experiments in which sodium salicylate or salicylic acid were applied to the intact rabbit skin in petrolatum USP XV or hydrated petrolatum USP XV provide further illustrations of the influence of the nature of the oint- ment on percutaneous absorption (161). 4-6 h after percutaneous applica- tion of petrolatum containing 6•o salicylic acid, a concentration of 6 mg 100 ml -• blood salicylic acid could be detected. When the hydrated ointment was used, the blood concentration reached a level of 9 mg 100 ml -• blood. On substituting the sodium salt for the acid in these preparations, a peak concentration of only 1 mg 100 ml -• could be detected in the blood. Addition of the surfactants, sorbitan monostearate, polyoxyethylene 20 or 40 sorbitan monostearate to petrolatum, enhanced the absorption of salicylic acid but did not affect that of sodium salicylate. Addition of these surfactants to the hydrophilic ointment reduced the absorption of salicylic acid but improved that of sodium salicylate. Salicylic acid is primarily an oil-soluble