Current Issue cover

Determining the Keratolytic Activities of Benzoyl Peroxide, Retinoic Acid and Salicylic Acid In vivo

Contact Author Howard I. Maibach, MD, University of California School of Medicine; Ali Alikhan, MD, Mayo Clinic; and Fatima S. Alikhan, Columbia University
Fill out my online form.

Keratolysis is the separation or loosening of the stratum corneum (SC), and is part of the natural cycle of skin renewal and regeneration. Disordered keratolysis, resulting in skin overgrowth or excessive desquamation, is responsible for various skin disorders. Keratolytics are agents that cause the separation of the SC through proteolysis of the desmosomes between keratinocytes,1, 3 and they therefore are often used to treat ailments that manifest in skin overgrowth. The use of keratolytics dates back to ancient Egypt when hydroxy acids such as lactic, citric and glycolic were commonly used for skin care.4 In fact, Cleopatra once relied on the keratolytic strength of lactic acid in milk to maintain the youthfulness appearance of her skin.4 Currently, keratolytics are employed to treat acne vulgaris, psoriasis, seborrheic dermatitis and warts, among other ailments. Specifically, patients afflicted with mild and severe acne utilize keratolytics such as benzoyl peroxide, retinoic acid and salicylic acid to slough off the outermost layers of excess and damaged skin.

Determining Keratolytic Efficacy

Keratolytic agents vary not only in strength, but also in the type of acne they are most effective in treating. In order to deduce the efficacy of an agent in removing superficial skin layers, many researchers conduct sequential adhesive tape stripping of a skin region and weigh the removed SC with precision balances. However, accurate measurements are time-consuming and may be distorted by water absorption and desorption from the tape or due to artifacts, e.g., applied vehicle and solutes in cases where topical agents are placed over the skin.5, 6 Thus, the authors instead propose evaluating various keratolytic agents by means of in vivo colorimetric protein assay, as described by Dreher et al.5

The colorimetric protein assay begins with the cutaneous application of an agent on a patch, which is then taped onto the subject’s skin for a predetermined number of hours. After this period, sequential tape stripping is performed on the site of topical treatment. The SC-adhering tapes are immersed and shaken in sodium hydroxide solution, resulting in extraction of the soluble SC protein fraction. This solution is then neutralized with HCl, as the assay is not effective under strongly alkaline conditions.

The protein assay is performed using a detergent compatible protein assay kita, which is similar to the Lowry biochemical assay7 for determining the amount of protein in a solution in that both are based on the reaction of protein—or more precisely, SC removed by tape-stripping—with an alkaline copper tartrate solution and folin reagent. The concentration of the reduced folin reagent is finally measured by light absorbance at a wavelength of 750 nm using a spectrophotometerb. This measurement allows for the quantification of microgram amounts of SC and diminishes the effects of confounding factors, e.g., moisture absorption and desorption, that plague simple tape weighing. The protein amount measured using this assay can be compared between various keratolytic agents, and statistical analyses allow for the determination of strong and weak keratolytics.


Using the described method, the authors evaluated the strength of three keratolytic agents—benzoyl peroxide (BPO), retinoic acid (RA) and salicylic acid (SA) (see Table 1). In a previously published study,8 the keratolytic efficacy of SA was determined as a function of pH. The test preparations were: an aqueous vehicle control at a pH of 7.4; a 2% SA aqueous solution with a pH of 3.3; a 2% SA aqueous solution at a pH of 3.3 with menthol; and a 2% SA aqueous solution with a pH of 6.95.

A statistically significant mass of SC was removed after 6 hr and 20 tape strips in all three experimental groups and compared with the vehicle, untreated and untreated but occluded groups. However, after just 10 strips, the pH 3.3 SA solution with menthol and the pH 6.95 SA solution removed significantly more SC than any other group, including the pH 3.3 SA solution. This data suggests that a neutral preparation of SC results in adequate SC uptake and a pronounced keratolytic effect. Moreover, the pH 6.95 SA solution was associated with the least skin irritation among the treatment groups. This finding differs from the findings by Leveque et al that demonstrated SA to be more effective in its acidic-pH, neutral form compared with its neutral-pH ionized form.9

In a second published keratolytic bioassay,1 SA, BPO and RA were examined. The test preparations were 0.05% all-trans RA, 2% SA at a pH of 6.95, 2% BPO, the vehicle, untreated skin, and occluded but untreated skin. After 3 hr of treatment, only the BPO treatment removed significantly more SC on 25 strips than untreated skin, while the other treatments did not achieve statistical significance. At 3 hr, SA had removed greater SC amounts on the first 10 superficial strips while deeper strips 11–25 demonstrated BPO to have the greatest SC removal. Statistically significant masses of SC were removed after 6 hr and 25 tape strips in all three experimental groups, compared with the vehicle, untreated and occluded groups. At 6 hr, the first 10 tape strips from the SA group removed more protein than the other groups; at 10–15 strips, treatments were comparable; and at 16–25 strips, protein removed from BPO sites was the greatest.

These in vivo human results indicate that all three treatments tested are effective keratolytics, which may account to some degree for their effectiveness against acne vulgaris. Furthermore, based on the stratified analysis of tape stripping, it appears that SA may be preferable for treating mild, superficial acne while BPO may be better suited for deeper, inflammatory acne lesions. The ability of BPO to loosen the SC at deeper levels complements its antimicrobial properties, resulting in an effective anti-inflammatory agent for papulopustular acne. Additionally, BPO appears to be effective even with short-term administration. RA showed inferior SC disruption at 3 hr but significant disruption at 6 hr, indicating time-dependent keratolytic effects, consistent with its well-studied interaction with nuclear receptors and alteration of gene transcription.

Future Directions

Etiological factors in acne vulgaris include excess sebum production, disordered follicular epithelial growth and differentiation, Propionibacterium acnes colonization of the hair follicle, and inflammation. Keratolytic agents act by normalizing disordered keratinization. The experiments detailed above allow for the precise quantification of keratolytic efficacy, which provides further scientific credibility to various acne treatments. The described experiments on BPO, SA and RA can be expanded to include other keratolytic acne agents as well. This knowledge is important to the creation of new products and formulations to treat acne vulgaris and optimizing exfoliating and other keratin-related therapies. Reproduction of the article without expressed consent is strictly prohibited.

1. JM Waller et al, “Keratolytic” properties of benzoyl peroxide and retinoic acid resemble salicylic acid in man, Skin Pharmacol Physiol 19 283–289 (2006)
2. Stedman’s Medical Dictionary for the Health Professions and Nursing, Lippincott Williams and Wilkins, Baltimore (2005)
3. S Hatakeyama et al, Retinoic acid disintegrated desmosomes and hemidesmosomes in stratified oral keratinocytes, J Oral Pathol Med 33 622–628 (2004)
4. L Baumann, Cosmetic Dermatology: Principles and Practice, McGraw-Hill Medical, New York (2009) pp 149
5. F Dreher, A Arens, JJ Hostynek, S Mudumba, J Ademola and HI Maibach, Colorimetric method for quantifying human stratum corneum removed by adhesive-tape stripping, Acta Derm Venereol 78 186–189 (1998)
6. E Martin, MT Neelissen-Subnel, FH De Haan and HE Bodde, A critical comparison of methods to quantify stratum corneum removed by tape stripping, Skin Pharmacol 9 69–77 (1996)
7. OH Lowry, NJ Rosebrough, AL Farr and RJ Randall, Protein measurement with the folin phenol reagent, J Biol Chem 193 1 265–275, PMID 14907713 (Nov 1951)
8. SJ Bashir et al, Cutaneous bioassay of salicylic acid as a keratolytic, Int J Pharmaceutics 292 187–194 (2005)
9. N Leveque, S Makki, J Hadgraft and P Humbert, Comparison of Franz cells and microdialysis for assessing salicylic acid penetration through human skin, Int J Pharm 269 323–328 (2004)

Related Content



Table 1. Studies using colorimetric protein assay to measure keratolytic potential

Table 1. Studies using colorimetric protein assay to measure keratolytic potential

Using the described method, the authors evaluated the strength of three keratolytic agents—benzoyl peroxide (BPO), retinoic acid (RA) and salicylic acid (SA) (see Table 1).

Footnotes (CT1107 Maibach)

a The Detergent Compatible Protein Assay Kit is manufactured by Bio-Rad Laboratories.
b The U-2001 UV-vis Spectrophotometer is manufactured by Hitachi.

Next image >