Post on 20-Jul-2020
transcript
Pharmacology and Therapeutics
Comparing 2.5%, 5%, and 10% Benzoyi Peroxide onInflammatory Acne Vulgaris
OTTO H. MILLS, JR., PH.D., ALBERT M . KLICMAN, M.D., PH.D., PETER POCHI, M.D.,
AND HARRIET COMITE, M.D.
From the Departments of Dermatology, University ofPennsy/vania School of Medicine, Philadelphia,Pennsylvania, and Boston University School of Medicine,Boston, Massachusetts
ABSTRACT: A 2.5% formulation of benzoyi peroxide wascompared with its vehicle, and with a 5% and a 10% proprietarybenzoyi peroxide gel preparation in three double-blind studiesinvolving 153 patients with mild to moderately severe acnevulgaris. The 2.5% benzoyi peroxide formulation was moreeffective than its vehicle and equivalent to the 5% and 10%concentrations in reducing the number of inflammatory lesions(papules and pustules). Desquamation, erythema, and symp-toms of burning with the 2.5% gel were less frequent thanwith the 10% preparation but equivalent to the 5% gel. The2.5% formulation also significantly reduced Propionibacteriumacnes and the percentage of free fatty acids in the surfacelipids after 2 weeks of topical application.
Benzoyi peroxide, in concentrations of 5%, 10%,and 20%, has been used effectively in the treat-
ment of acne for more than 20 years.''* This compoundhas been shown to suppress Propionibacterium acnesin vivo, the probable basis for its therapeutic effect.*With such concentrations, side effects such as ery-thema, desquamation, and burning, itching, or stingingare fairly common. This paper describes clinical trialsofa 2.5% benzoyi peroxide gel, which was comparedwith 5% and 10% benzoyi peroxide gels in groups ofpatients with inflarnmatory acne vulgaris. Antibacterialand jipid studies were also performed on the 2.5%benzoyi peroxide formulation.
Address for correspondence: Otto H. Mills, Jr., Ph.D., UMDNJ-Robert Wood Johnson Medical School, One Robert Wood JohnsonPlace, New Brunswick, N] 08903.
Subjects, Materials, and Methods
Clinical Studies
The same methods were used to conduct three double-blind studies. After giving informed consent, subjects withmild to moderately severe inflammatory acne vulgaris of theface (minimum of 10 inflammatory lesions) were assigned toone of the three treatment groups. A total of 153 subjects,74 men and 79 women (average age of 20 years), participatedin the three studies. In the first, 25 subjects used 2.5% ben-zoyi peroxide gel and 25, the gel vehicle for this formulation.In the second, 26 used the 2.5 gel and 27, a 5% gel. Thethird study consisted of 25 subjects who used the 2.5% geland 25, a 10% benzoyi peroxide gel. The subjects receivedno medications for any reasor\s during \he 4-week periodprior to the start of the study.
The study participants were instructed to wash daily witha non-medicated soap, rinse, and dry with a clean towelbefore applying the study medication to the face twice daily(morning and evening) for 8 weeks. Subjects were examinedbefore treatment for baseline determinations and at weeks2, 4, 6, and 8 after the start of treatment. At each visit, thenumber of facial inflammatory lesions (papules and pustules)was recorded. In addition, the frequency and severity of sideeffects such as erythema, peeling, and burning were noted,A global assessment of improvement was also made by theinvestigator at each visit, according to the following criteria:"excellent," greater than 75% improvement; "good," about50% improvement; "fair," about 25% improvement; and"poor," little or no improvement.
Antibacterial and Lipid Studies
Ten subjects who had a P. acnes count of 100,000 coloniesper cm^ or greater on the face were selected for this study.A score of 3 or greater on the follicular porphyrin fluorescencescale was also required for admission. The density of P. acnes,the degree of porphyrin fluorescence, and the ratio of freefatty acids to triglycerides found in Upid samples were de-termined before and after 7 and 14 days of twice-daily ap-plications ofthe 2.5% benzoyi peroxide formulation to theface. All applications were made by laboratory technicians.
At each sample day (0, 7, 14), the skin was prepared bywiping the surface of the forehead for 30 seconds with apiece of gauze saturated with 0.1 % Triton X-100 (Rohm andHaas, Philadelphia, PA), followed by a distilled water rinse
664
No. 10 BENZOYL PEROXIDE IN ACNE Mills et al. 665
TABLE 1. Comparison of Effects of 2.5% Benzoyi Peroxide and Vehicle
Week
02468
Number ofSubjects
2525242525
2.5% Benzoyi Peroxide
Mean Numberof Lesions*
13.610.89.37.56.8
Mean %Reduction
20.4%31.4%44.10/050.9%
Number ofSubjects
2525252225
Vehicle
Mean Numberof Lesions*
13.713.813.013.111.2
Mean %Reduction
-2.70/0
2.8%3.40/0
17.60/0
P Valuet for
AverageTreatmentDifference
0.9U<0.01
0.02<0.01
0.01
* Papules and pustules,t By t-test.
Difference in baseline counts.
and a 30-second wipe with hexane. This procedure removedenvironmental contaminants, desquamating cells, and surfacebacteria and lipids. Two areas were protected by plasticweighing boats with several perforations to allow evaporationof sweat. After 1 hour, the site was sampled by the detergentscrub technique of Williamson and Kligman.' A sterile glasscylir^der with an interna) area of 3.8 cm^ was placed over thesite. One ml of 0.1% Triton X-100 in 0.075% phosphatebuffer (pH 7.9) was added and the surface scrubbed with ablunt Teflon (Arthur H. Thomas, Philadelphia, PA) spatulafor one minute. This procedure was repeated, and the twosamples were pooled. Subsequently, tenfold dilutions weremade in 0.5% buffered Triton X-100; the samples were drop-plated on brain heart agar with 0.1 % Tween 80 (Atlas Chem-ical, Wilmington, DE) and incubated anaerobically for 7 daysin a Gas Pak (Arthur H. Thomas, Philadelphia, PA) jar system.
P. acnes was identified by colony morphology, by suscep-tibility to P. acnes bacteriophage, and, when indicated, bybiochemical testing.'""
At the other site on the forehead, lipid samples were col-(ected by putting 2 ml of hexane-containing methyl nervon-ate, as an internal standard, in the glass cup, as above. Thesite was scrubbed for 30 seconds with a blunted Teflon po-Jjceman. The solution was taken up on a Pasteur pipette andpassed through a 0.45-millipore filter to remove skin debrisand bacteria. It was then placed in Teflon-capped glassscrewtop vials. The vials were uncapped, dried overnight ina vacuum at 40 C, then capped and stored at 40 C until lipidthin-layer chromatography was done by the method ofDowning.'^
Each subject was examined under Wood's light for por-phyrin fluorescence prior to washing and obtaining samples."
The intensity of fluorescence was graded on a 0 to 6 scalewith 0 = none, 1 to 2 = mild fluorescence, 3 to 4 = moderate,and 5 to 6 = heavy fluorescence.
Statistical Methods
Fisher's exact test" was used to compare treatment groupswith respect to side effects at each visit during the treatmentperiod. A group t-test'" was used to compare treatmentgroups with regard to the reduction in number of papulesand pustules from baseline. A paired t-tesV* was used toanalyze changes from baseline counts within treatmentgroups. The Wilcoxon rank-sum test"* was used to analyzechanges estimated by global ratings.
Results
Results on efficacy and side effect data are presentedseparately for each study. No subject was obliged todrop out of any study because of adverse effects.
Study 1: 2.5% Benzoyi Peroxide Versus Vehicle
The 2.5% benzoyi peroxide was more effective thanthe vehicle in reducing the number of inflammatorylesions (papules and pustules) at all follow-up visits(Table 1).
The 2.5% benzoyi peroxide was also significantlymore effective than the vehicle in global ratings at allevaluations. Mild peeling, burning, and itching were
TABLE 2. Comparison of Effect of 2.5% and 5.0% Benzoyi Peroxide Cel
2.50/0 Benzoyi Peroxide 50/0 Benzoyi Peroxide P Valuet for
Number ofWeek Subjects
02468
*
t
2626252626
Papules and pustules.By Cutest.
Mean Numberof Lesions*
21.314.813.310.09.6
Mean 0/0Reduction
32.2%40.30/0
54.30/0
55.90/0
Number ofSubjects
2727272525
i Difference
Mean Numberof Lesions
19.413.512.910.6
7.8
in baseline counts.
Mean %Reduction
30.60/0
35.10/0
47.20/0
57.70A,
AverageTreatnientDifference
0.47^:0.750.510.410.94
666 INTERNATIONAL JOURNAL OF DERMATOLOGY December 1986 Vol. 25
TABLE 3. Comparison of Effects of 2.5% and 10% Benzoyi Peroxide Cel
Week
02468
Number ofSubjects
2525242424
2.5% Benzoyi Peroxide
Mean Numberof Lesions*
19.716.112.810.910.5
Mean %Reduction
18.3%35.0%44.7%46.7%
Number ofSubjects
2425242324
10% Benzoyi Peroxide
Mean Numberof Lesions
23.719.014.914.513.2
Mean %Reduction
19.8%37.-1%38.8%44.7%
P Valuet for
AverageTreatmentDifference
.17*
.66
.47
.75
.57
* Papules and pustules,t By t-test.
more frequent in the benzoyi peroxide group than inthe vehicle group, but only statistically significantly sofor peeling at week 8. There was no significant differ-ence in the incidence of erythema, although 2.5%benzoyi peroxide more often induced erythema atweek 2.
Study 2: 2.5% Versus 5.0% Benzoyi Peroxide
No significant difference in efficacy between the2.5% and the 5.0% benzoyi peroxide formulations wasnoted. In both groups, a significant reduction in pa-pules and pustules was observed at 2,4, 6, and 8 weeks(Table 2). The global ratings confirmed the lack of sig-nificant difference in efficacy between the 2.5% and5.0% gel formulations. There was no significant dif-ference between the two preparations in regard toburning, peeling, or erythema.
Study 3: 2.5% Versus 10% Benzoyi Peroxide
Both 2.5% and 10% benzoyi peroxide gels reducedthe number of papules and pustules from baselinecounts, but there was no statistically significant differ-ence between the two groups (Table 3). Statisticalevaluation ofthe investigator's global response ratingsalso indicated that there was no significant differencebetween the efficacy of 2.5% and 10% benzoyi per-oxide.
There was a statistically significant difference in thefrequency and severity of burning, erythema, andpeeling among subjects who used 10% benzoyi per-oxide than among those who used the 2.5% concen-tration at all follow-up visits (Table 4).
Bacteriology and Free Fatty Acids
A marked reduction in the quantity of P. acnes wasobserved after 1 week (Table 5). The intensity of fol-licular porphyrin fluorescence was also reduced by 1week and markedly suppressed by 2 weeks. There was
Difference in baseline counts.
also a significant reduction in the ratio of free fattyacids to triglycerides.
Discussion
The 2.5% benzoyi peroxide formulation was signif-icantly more effective than its vehicle in reducing thenumber of papules and pustules and was comparableto the 10% benzoyi peroxide by lesion counts. By thesame measurement, there were no differences be-tween the 2.5% gel and the 5% benzoyi peroxide gel;both were clinically effective. The incidence of irrita-tion was lower with 2.5% than with 10% benzoyi per-oxide. It should be pointed out that in two of theseclinical studies there would need to have been muchlarger patient groups to assure "statistical power" fordifferences between treatments. The differences be-tween the 2.5% benzoyi peroxide and its vehicle isnot a question. A clear significant difference betweenthese two exists. When the 2.5% versus 5% and 2.5%versus 10% studies were reviewed for "statisticalpower," it is evident that, with the number of subjectsinvolved, the power of the test was not high enoughto assure a difference that was statistically significant.However, we feel these studies are clinically significantand present important information for clinicians andthose working in dermatopharmacology.
Also, these studies do not represent a titration ofpercent concentration of drug in the same vehicle. The2.5% formulation vehicle was different from those of
TABLE 4. Frequency and Severity of Burning, Erythema, andPeeling* (Total Number of Reports for 8 Weeks)
2.5% Benzoyi Peroxide Cel
BurningErythemaPeeling
Mild
202250
Moderate
149
10% Benzoyi Peroxide Gel
BurningErythemaPeeling
Mild
575136
Moderate
203055
* Possibly or probably relaVed to drug.
No. 10 BENZOYL PEROXIDE IN ACNE Mills et ai. 667
TABLE S. Effect of Topical Application of 2.5% Benzoyi Peroxide Cel on Quantitative P. acnes Counts,Follicular Porphyrin fluorescence and Free Fatty Acids in Skin Surface Lipids
SubjectNumber
123456789
10Meanp value
WeekO
5.82856.27755.27756.64556.10155.62436.73836.36476.43725.70356.083
—
P. acnes (log/cm^)
Week 1
4.92544.36473.75474.02235.55383.16845.66574.02235.73834.60214.504
<0.001
Week 2
4.46943.64553.64554.49944.02233.40255.60213.73835.64553.16844.108
<0.001
Follicular
WeekO
66654456565.30—
Porphyrin Fluorescence(Grades 0-6)
Week 1
45643434344.00
<0.01
Week 2
23422213222.30
<0.001
Free
WeekO
0.470.840.380.221.770.931.250.221.050.190.732
—
Fatty Acids/Triglycerides
Week 1
0.160.730.160.191.120.861.190.160.990.190.575
<0.02
Week 2
0.140.660.140.110.660.550.890.140.600.100.419
<0.01
the 5% and 10% formulations, but we think it impor-tant that we saw these results with percent concentra-tions of drug that were one-fourth to one-half less thanthe highest concentration.
The laboratory results from the in vivo study of P.acnes showed the 2.5% benzoyi peroxide gel reducedthe anerobic population by 97% after twice-dailytreatments for 1 week and by 99% after 2 weeks. Thisoutcome is in agreement with previous work using adifferent 2.5% benzoyi peroxide formulation.'^
Regarding the clinical changes of peeling and ery-thema and the symptoms of burning, there were nodifferences between 2.5% versus 5%, but differencesdid exist between the 2.5% and its vehicle and the2.5% and 10% formulations. With this in mind, thelower concentration of benzoyi peroxide should beuseful for treating patients with easily irritated skin.Also, in combination topical therapy with comedoiyticagents, 2.5% benzoyi peroxide might lessen the ex-pected degree of irritation.
Drug Names
2.5% benzoyi peroxide: Clear by Design5.0% benzoyi peroxide: Desquam X-510.0% benzoyi peroxide: Desquam X-10
References
1. Frank L. Active oxygen acne therapy—oxygenation vs. reductionofthe follicular structures. Cutis. 1965;1:306-308.
2. Pace WE. A benzoyi peroxide-sulfur cream for acne vulgaris.Can Med Assoc ). 1965;93:252-254.
3. Danto )L, Maddin WS, Steward WD, et al. A controlled trial ofbenzoyi peroxide and precipitated sulfur cream in acne vul-garis. AppI Ther. 1966,8:624-625.
4. Belknap BS. Treatment of acne with 5 percent benzoyi peroxidegel or 0.05 percent of retinoic acid cream. Cutis. 1979;23:856-859.
5. Smith EB, Padilla RS, McCabe )M, et al. Benzoyi peroxide lotion(20 percent) in acne. Cutis. 1980,25:90-92.
6. Kligman AM, Mills OH, McCinley K|, et al. Acne therapy withtretinoin in combination with antibiotics. Acta Derm Verereoi.1975,74(Suppl):111-115.
7. Martin R), Kahn G, Gooding JW, et al. Cutaneous porphyrinfluorescence as an indication of antibiotic absorption and ef-fectiveness. Cutis. 1973;12:758-764.
8. McGinley KJ, Webster GF, Leyden JJ. Facial follicular porphyrinfluorescence: correlation with age and density of Propioni-bacterium acnes. BrJ Dermatol. 1980;102:437-441.
9. Williamson P, Kligman AM. A new method for the quantitativeinvestigation of cutaneous bacteria. J Invest Dermatol.1965,45:498-503.
10. Marples RR, McGinley KJ. Corynebacterium acnes and the an-aerobic diphtheroids from human skin. J Med Microbiol.1974;7:349-357.
11. McGinley KJ, Webster GF, Leyden JJ. Regional variation of cu-taneous propionibacterium. J AppI Env Microbiol. 1978;35:62-66.
12. Downing DT. Photodensitometry in the thin-layer chromato-graphic analysis of neutral lipids. J Chromatogr. 1968;38:91-99.
13. Ostle B. Statistical inference: testing hypotheses: statistics in re-search. Iowa City, IA: Iowa State University Press, 1963;119-321.
14. Seigel S. The case of k independent samples: nonparametricstatistics for the behavioral sciences. New York: McGraw-Hill, 1956:101-117.
15. Leyden JJ, McGinley KJ, Mills OH, et al. Topical antibiotics andtopical antimicrobial agents in acne therapy. Acta Derm Ve-nereol. 1980;89(Suppl):75-82.