Dibenzoyl peroxide

Carcinogenicity

Description of key information

Two-year dermal carcinogenicity studies conducted according to GLP found no evidence of oncongenicity resulting from daily topical application of a benzoyl peroxide gel in rats and mice at doses that meet the maximum tolerance dose. The other data on the carcinogenicity potential of exposure to benzoyl peroxide has been largely reviewed (Kraus, 1994; UNEP, 2002; IARC, 1985, 1999) and the conclusions were somewhat similar: despite a long-term human exposure, and a wide use, no carcinogenic effect was reported in human; tumor promotion in mouse skin occurs only under specific experimental conditions and predominantly in highly sensitive strains. Furthermore, there is significant physiological differences between mouse and human skin that does not allow to extrapolate directly findings to humans. Otherwise  clinical uses scenarios are different from the conditions resulting in tumor promotion in mouse skin.
In conclusion, there is no evidence to suggest that benzoyl peroxide is a carcinogen. However, there is some evidence from non-guidelines studies that benzoyl peroxide is a skin tumor promoter. Based upon lack of carcinogenic effects in more than 23 animals carcinogenicity studies, on negative findings in epidemiological studies, and on the absence of carcinogenic potential for the main metabolite (benzoic acid), there is no toxicological concern for benzoyl peroxide carcinogenicity in humans. According to IARC evaluation (1999), there is inadequate evidence in humans for the carcinogenicity of benzoyl peroxide. There is limited evidence in experimental animals for the carcinogenicity of benzoyl peroxide. Overall evaluation: Benzoyl peroxide is not classifiable as to its carcinogenicity to humans (Group 3).

Key value for chemical safety assessment

Carcinogenicity: via oral route

Link to relevant study records

Reference

Endpoint:

carcinogenicity: oral

Type of information:

experimental study

Adequacy of study:

weight of evidence

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

other: Although it may not meet all aspects of current standards for carcinogenicity testing, this study was carefully conducted and the mortality rates were generally quite low for studies of their time.

Principles of method if other than guideline:

Feeding toxicity and carcinogenicity study

GLP compliance:

no

Species:

rat

Strain:

other: Albino

Sex:

male/female

Details on test animals or test system and environmental conditions:

Diet composition: 100% wholemeal flour, 56; dried skimmed milk, 20; fish meal, 11; alfalfa meal, 2; dried yeast and yeast extract, 1 each; cod liver oil, 3; salt mixture, 4. The remaining 2% was a triturate in whole meal flour of benzoyl peroxide added in the form of Novadelox, a commercial powder containing 18% benzoyl peroxide, 78% calcium sulphate and 4% magnesium carbonate.

Route of administration:

oral: feed

Analytical verification of doses or concentrations:

not specified

Duration of treatment / exposure:

120 weeks

Frequency of treatment:

ad libitum

Post exposure period:

none

Remarks:

Doses / Concentrations:
0.0157, 0.157 and 1.57 % of Novadelox in diet
Basis:
nominal in diet
groups F3, F2 and F1, respectively

Remarks:

Doses / Concentrations:
28, 280, 2,800 mg BPO/kg diet
Basis:
nominal in diet
groups F3, F2 and F1, respectively

Remarks:

Doses / Concentrations:
Male: 1.1, 11.2 and 112 mg/kg bw/d
Basis:
other: based on food consumption of 40 g/kg bw/d

Remarks:

Doses / Concentrations:
Female: 1.4, 14 and 140 mg/kg bw/d
Basis:
other: based on food consumption of 50 g/kg bw/d

No. of animals per sex per dose:

25

Control animals:

other: yes, concurrent no treatment (FC group)

Observations and examinations performed and frequency:

Body wt gains of rats were noted weekly for 18 weeks, and then monthly. The animals were examined twice daily with regard to their general health. Animals that were moribund were killed. All animals that were killed or that died were autopsied. Benign and malignant tumours were analysed.

Clinical signs:

no effects observed

Mortality:

no mortality observed

Body weight and weight changes:

effects observed, treatment-related

Food consumption and compound intake (if feeding study):

not specified

Food efficiency:

not examined

Water consumption and compound intake (if drinking water study):

not examined

Ophthalmological findings:

not examined

Haematological findings:

not examined

Clinical biochemistry findings:

not examined

Urinalysis findings:

not examined

Behaviour (functional findings):

not examined

Organ weight findings including organ / body weight ratios:

not examined

Gross pathological findings:

not specified

Histopathological findings: non-neoplastic:

effects observed, treatment-related

Histopathological findings: neoplastic:

no effects observed

Details on results:

CLINICAL SIGNS AND MORTALITY
Some statistically significant differences were found between the mortality rates of control and experimental groups, but the importance of these findings is doubtful. The mortality of female rats fed on diet F2 was greater than the controls (group FC) at 78 weeks, whereas before and after this period no significant difference was found. There was no increased mortality in males receiving this diet or in males and females receiving the diet containing a higher concentration of Novadelox.

BODY WEIGHT AND WEIGHT GAIN
The rate of body weight gain was depressed, in several cases to a significant extent, in male and female rats consuming the flour diets containing the two highest levels of Novadelox. The effect, which was most noticeable during the first 8 months with diet F1, could have been due to (1) a diminished acceptability of the diet to rats, (2) a marginal nutritional deficiency caused by the action of benzoyl peroxide on flour, or to (3) the toxic action of
Novadelox or its reaction products.

HISTOPATHOLOGY: NON-NEOPLASTIC
Many and varied gross and microscopic changes were found in all groups. Except for testicular atrophy, pathological abnormalities occurred to an equal extent in experimental and control animals or, because of a low incidence, occurred randomly in control or experimental groups. Infective changes were common, particularly in the lungs of both sexes of rat, and in the uterus of female rats. More than one lesion caused by infection was found per animal in rats. Local vascular damage frequently accompanied infective lesions. Degenerative changes were found most frequently in the liver, kidney, testes and adrenals of rats. Ulceration of the stomach was not found in control rats, while it occurred in 2 of the 96 examined which received the highest dosage level of benzoyl peroxide in flour diet.

HISTOPATHOLOGY: NEOPLASTIC (if applicable)
The overall tumour incidence was not significantly different between the experimental and control groups in rats. Most tumoun were found in rats which were examined after 100 weeks treatment, the occurrence being low before that time.

Relevance of carcinogenic effects / potential:

Dibenzoyl peroxide didn't elicit any signf of carcinogenic activity in rats.

Dose descriptor:

NOEL

Effect level:

> 112 - 140 mg/kg bw/day

Sex:

male/female

Basis for effect level:

other: No carcinogenic effect.

Remarks on result:

other: Effect type: carcinogenicity (migrated information)

Dose descriptor:

NOAEC

Effect level:

11.2 - 14 mg/kg bw/day

Sex:

male/female

Remarks on result:

other: Effect type: toxicity (migrated information)

Dose descriptor:

LOAEL

Effect level:

112 - 140 mg/kg bw/day

Sex:

male/female

Basis for effect level:

other: Decreased body weight gain and testicular atrophy

Remarks on result:

other: Effect type: toxicity (migrated information)

Conclusions:

The overall tumor incidence was not significantly different between the experimental and control groups.

Executive summary:

Sharratt et al. conducted a study on the treatment of whole meal flour with benzoyl peroxide. Albino rats (groups of 25 animals/sex/ species; strain and age were unspecified) were fed a diet with increasing amount of Novadelox, a commercial powder containing 18% benzoyl peroxide (purity unspecified). Control groups received a diet lacking Novadelox. According to the authors, treatment groups received 10, 100 or 1,000 times the daily intake, for 120 weeks. The estimated doses of benzoyl peroxide amounted to 28, 280 and 2,800 mg/kg diet. Body weight gain was depressed in male and female rats consuming the diets containing the highest levels of benzoyl peroxide. A part from the 280 mg/kg diet group, in which there was a large number of accidental deaths, there was no significant difference between experimental and appropriate control groups in the mortality rate of the animals. No evidence for benzoyl peroxide-related carcinogenicity was observed.

Endpoint conclusion

Endpoint conclusion:

no adverse effect observed

Study duration:

chronic

Species:

other: rat, mouse

Carcinogenicity: via inhalation route

Endpoint conclusion

Endpoint conclusion:

no study available

Carcinogenicity: via dermal route

Link to relevant study records

Reference

Endpoint:

carcinogenicity: dermal

Type of information:

experimental study

Adequacy of study:

weight of evidence

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

other: GLP guideline study

Qualifier:

equivalent or similar to guideline

Guideline:

OECD Guideline 451 (Carcinogenicity Studies)

Deviations:

no

GLP compliance:

yes

Species:

rat

Strain:

Fischer 344

Sex:

male/female

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Charles River Laboratories, Inc., Portage, Michigan.
- Age at study initiation: 6-week-old,
- Weight at study initiation: from 94 to 130 g for the males and 80 to 106 g for the females
- Housing: individually in stainless-steel, hanging, wire-mesh cages measuring 24.2 x 22.0 x 17.3 cm (d x w x h)
- Diet: PMI® Certified Rodent Diet® #5002 was available ad libitum
- Water: tap water, ad libitum
- Acclimation period: 2 weeks

ENVIRONMENTAL CONDITIONS
- Temperature (°C): 22 ± 3
- Humidity (%): 55± 15
- Air changes (per hr): > 10
- Photoperiod (hrs dark / hrs light): 12/12

Route of administration:

dermal

Vehicle:

other: aqueous carbopol gel

Details on exposure:

TEST SITE
- Area of exposure: 3.5x5 cm of dorsal skin at the intraseapular region
- Type of wrap if used: none used
- Time intervals for shavings or clipplings: At least 24 hours prior to the first dose, and weekly thereafter, as needed, an area of dorsal skin that was
larger than but included the application site was clipped free of hair to allow uniform application of doses and clear observation of the application site
- Occlusion: no

REMOVAL OF TEST SUBSTANCE
- Washing (if done): no

TEST MATERIAL
- Amount(s) applied (volume or weight with unit): 0.3 mL
- Concentration in vehicle: 1.67, 5, and 15 percent BP
- Constant volume or concentration used: yes

VEHICLE
- Justification for use and choice of vehicle: in a preliminary test, the influence of vehicle on skin response to benzoyl peroxide was evaluated in F344 rats. Two preparation were tested: BPO in acetone or suspension in carbopol gel. The macroscopic and microscopic skin effects were compared in 14 day studies. There were no difference in results, therefore the carbopol gel was selected as the vehicle considering its relevance with respect to human relevance. (Binder et al., 1997)

USE OF RESTRAINERS FOR PREVENTING INGESTION: no

Binder et al., 1997: Influence of vehicle on skin response to benzoyl peroxide (BPO) in F344 rats and B6C3F1 mice, Toxicologist 36 188 1997

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

U.S.P. Official Monographs XXII method. Every 6 months, based on the study timellne, analyses were conducted on all concentration batches used or in use during the previous 6 months.

Duration of treatment / exposure:

at least 104 weeks

Frequency of treatment:

7 days a week

Post exposure period:

50 male and female animals were treated with high-dose test article for 52 weeks and then treated with vehicle until terminal sacrifice.

Remarks:

Doses / Concentrations:
1.67, 5.0 and 15.0 %
Basis:
nominal conc.

Remarks:

Doses / Concentrations:
5, 15 and 45 mg/rat/day
Basis:
other: nominal dose level (volume 0.3 ml/rat)

Remarks:

Doses / Concentrations:
0.3, 0.8 and 2.6 mg/cm²
Basis:
other: nominal dose level (surface 17.5 cm²)

No. of animals per sex per dose:

60 rats/sex/group for nearly all the groups.

Control animals:

yes, concurrent no treatment

yes, concurrent vehicle

other: sentinel group

Details on study design:

Dosage Level Concentration N Number of Animals
Group mg/rat/d % BPb Males Females
1 (Vehicle Control) 0 0 60a 60a
2 (Low) 5 1.67 60a 60a
3 (Mid) 15 5.0 60a 60a
4(High)e 45 15.0 60a 60a
5 (High Discontinued)c 45 15.0 50 50
6 (Untreated Control) 60a 60a
7 (Sentinels)d 20 20

a The first 10 animals/sex/group were designated as Interim Sacrifice mice. and were sacrificed after 52 weeks of treatment.
b Percent benzoyl peroxide in carbopol gel, corrected for water content.
c Animals were treated with high-dose test article for 52 weeks and then treated with vehicle until terminal sacrifice.
d Sentinel animals did not receive treatment. Blood was collected from five predesignated animals/sex during Weeks 26, 52, 78 and 104. Sera were used for pathogen screening.
e Group 4 animals received 25% BP during weeks 1-56: vehide during Weeks 57-58, 85-86 and 93-105 and 15% BP during Weeks 59-84 and 87-92.

Positive control:

No

Observations and examinations performed and frequency:

CLINICAL OBSERVATIONS
The rats were observed twice daily (a.m. and p.m., at least 6 hours between observations) for evidence of mortality and moribundity. Once weekly, each animal was removed from its cage and examined for abnormalities and signs of toxicity, specifically noting the location, size, and appearance of any grossly visible or palpable masses. The treated skin for analogous site on the untreated control) was graded for irritation once weekly.

BODY WEIGHT
Body weights were recorded weekly from weeks 1 through 14 and every fourth week thereafter and at weeks 53 and 105.

FOOD CONSUMPTION
Food consumption was measured and recorded weekly for weeks 1 through 13 and every fourth week thereafter and at weeks 52 and 104.

HAMEATOLOGY
Blood smears were prepared from all moribund-, interim-, and terminal sacrifice animals for possible evaluation of hematopoietic neoplasia.

Sacrifice and pathology:

GROSS PATHOLOGY: Yes
All animals which were found dead or sacrificed in extremis during the study were subjected to a gross postmortem examination. The first 10 surviving predesignated rats/sex in Groups 1, 2,3,4, and 6 were sacrificed following 52 weeks of treatment. The remaining animals were sacrificed during Weeks 105 and 106. At necropsy, macroscopic observations were recorded, and selected tissues were collected and preserved.

ORGAN WEIGHT
The liver with gallbladder, kidneys, and brain were weighed from all animals at interim sacrifice.

HISTOPATHOLOGY: Yes (see table) / No / No data
Histologic evaluations were conducted on the livers and skin (treated and untreated) from all Group 1 through 4 and Group 6 mice sacrificed after 52 weeks of exposure and on the skin (treated and untreated) from all Group 5 mice. All tissues listed were evaluated histologically from all Group 1 through 4 and Group 6 mice sacrified after 104 weeks of exposure and from unscheduled deaths.

Tissues list: Adrenals, aorta (thoracic), bone with marrow (sternum, femur), brain with brainstem (medulla/pons, cerebellar cortex, and cerebral cortex), cervical spinal cord, clitoral gland, cecum, colon, rectum duodenum, jejunum, ileum, esophagus, eyes (with optic nerve and contiguous harderian gland), femur including articular surface, heart, kidneys, lacrimal gland (exorbital), larynx, lesions, liver (collected whole, left lateral, right lateral and median lobes examined microscopically), lumbar spinal cord, lungs with bronchi, lymph nodes (mandibular, mesenteric, and regional when present in area draining a mass), mammary gland with skin, mid-thoracic spinal cord, nasal cavity and nasal turbinates, ovaries, pancreas, pituitary, preputial gland, prostate, salivary glands (mandibular), sciatic nerve, seminal vesicles, skeletal muscle (thigh), skin (treated and untreated), spleen, sternum, stomach (glandular and nonglandular), testes with epididymides, thymus, thyroid with parathyroids, tissue masses, tongue, trachea, ureters, urinary bladder, uterus with vagina and cervix, Zymbal' s gland (auditory sebaceous gland).

Clinical signs:

no effects observed

Mortality:

no mortality observed

Body weight and weight changes:

no effects observed

Food consumption and compound intake (if feeding study):

no effects observed

Food efficiency:

not examined

Water consumption and compound intake (if drinking water study):

not examined

Ophthalmological findings:

not examined

Haematological findings:

not examined

Clinical biochemistry findings:

not examined

Urinalysis findings:

not examined

Behaviour (functional findings):

not examined

Organ weight findings including organ / body weight ratios:

no effects observed

Gross pathological findings:

effects observed, treatment-related

Description (incidence and severity):

Skin irritation at all dose levels

Histopathological findings: non-neoplastic:

effects observed, treatment-related

Description (incidence and severity):

Skin irritation at all dose levels

Histopathological findings: neoplastic:

no effects observed

Details on results:

CLINICAL SIGNS AND MORTALITY
- Survival
There were no treatment-related effects on survival. Statistical evaluation of survival data revealed no significant positive or negative trend in the male or female mortality, nor were there significant group comparisons observed in either sex. Survival rates through Week 52 (prior to the Week 53 interim sacrifice) were 100% for Group 1,3,4, and 5 males, 98% for Group 2 and 6 males, 100% for Group 1-4 and 6 females, and 94% for Group 5 females. Group 1-6 survival rates through Week 104 were 74, 75, 78, 80, 86, and 74% for the males and 84, 78, 88, 88, 72, and 78% for the females, respectively.
- Clinical Observations
There were no treatment-related clinical observations. In general, the findings observed occurred sporadically and/or were of the type commonly seen in this species at this
laboratory.
- Application Site Observations
Dermal Irritation: Incidences of slight erythema were seen in both treated and vehicle control animals. Moderate erythema was noted only in two rats (one male and one female) that received 15% benzoyl peroxide. However, due to the low incidence (2 of 120 rats), this finding cannot be definitely attributed to treatment. The incidence and extent of test material accumulation at the application site was dose-related.
Dermal Masses: Masses at the application site were observed in males only. Masses were noted only in the Group 4 (continuous treatment with 15% benzoyl peroxide), Group 5 (discontinuous 15% benzoyl peroxide), and Group 6 (negative control) animals. These findings are not considered treatmentrelated since the negative control (untreated) animals exhibited the greatest incidence of findings .

BODY WEIGHT AND WEIGHT GAIN
There were no treatment-related differences in body weights and/or body weight changes.

FOOD CONSUMPTION
There were no treatment-related changes in food consumption.

ORGAN WEIGHTS
Mean terminal body weight, absolute organ weight, organ-to-body weight percentage, and organ-to-brain weight ratio values for the interim-sacrifice animals were similar among all treatment groups.

GROSS PATHOLOGY
- Unscheduled Deaths
Gross pathology findings for animals with unscheduled deaths included but were not limited to, indented ventral surface of the brain, enlarged pituitary gland, enlarged spleen, granular/pitted/rough liver, dark kidneys, and dark areas on the stomach for both males and females. Gross pathology findings for male animals also included unequally sized, small, soft testis filled with dark or pale material, and small seminal vesicles. None of these findings occurred with any apparent pattern and are not considered treatment-related.

- Interim Sacrifice
There were no treatment-related necropsy findings noted at the interim sacrifice.

- Terminal Sacrifice
There were no treatment-related necropsy findings noted at the terminal sacrifice.

HISTOPATHOLOGY: NON-NEOPLASTIC
- Unscheduled Deaths
The most commonly cited causes of early death, including animals that were sacrificed in extremis, were hematopoietic neoplasia (mostly mononuclear cell leukemia) and pituitary neoplasia. Both findings are prevalent in normal populations of F-344 rats as they age, and both were randomly scattered among exposure groups within this study.

- Interim Sacrifice
Microscopic evaluation revealed that the only treatment-related histopathologic findings noted were those observed at the application site. These treatment-related skin findings consisted of hyperkeratosis, acanthosis, sebaceous gland hyperplasia, and chronic subepidermal inflammation. These findings were observed in male and female test article treatment groups (Groups 2,3 and 4) with the severity of findings increased at the higher dose levels. Changes were mostly of a minimal to mild degree of severity. Evidence of hyperkeratosis was observed in a few vehicle control (Group 1) and untreated control (Group 6) animals. Histopathologic changes observed in the kidneys were all of the kinds usually encountered in normal populations of rats of this strain and age and were not related to treatment.

- Terminal Sacrifice
Aside from expected increases in normally occurring age-related background changes, the findings after 104 weeks of exposure were similar to those that occurred after 52 weeks. Microscopic evaluation revealed that the only treatment-related histopathologic findings noted were those seen at the application site. These consisted of hyperkeratosis, acanthosis, sebaceous gland hyperplasia, and chronic subepidermal inflammation. As with the interim-sacrifice animals, these changes occurred in male and female test article treatment groups (Groups 2, 3 and 4) and were mostly of a minimal to mild degree of severity. Also consistent with the interim-sacrifice animals, hyperkeratosis was noted in a few vehicle control (Group 1) and untreated
control (Group 6) animals. These changes progressed only slightly between 52 and 104 weeks of exposure.
For the Group 5 animals (52 weeks of treatment with 15% benzoyl peroxide followed by 52 weeks of recovery/vehicle), hyperkeratosis was present approximately twice as frequently as it was in either control group (Groups 1 and 6), but less than half as frequently as in any other group, including that group exposed to the lowest level (1.67% benzoyl peroxide) for either 52 or 104 weeks. Acanthosis was not appreciably increased in the recovery group compared to controls, and the finding of sebaceous gland hyperplasia in Group 5 was equivocal. These results, in general, imply good. although
not complete, recovery from the effects of 52 weeks of treatment at the high-dose level with remnants of hyperkeratosis remaining after 52 weeks of recovery.

HISTOPATHOLOGY: NEOPLASTIC
There were no histologic findings suggestive of any toxic or oncogenic effects from exposure to benzoyl peroxide gel. Although some of the neoplastic findings (e.g. large granular lymphocytic leukemia in females) showed borderline significance at the p < 0.05 level, they were of inconsistent nature (i.e., nonmonotonic) characteristic of aging F-344 rats, and do not suggest a carcinogenic response. Also, in the specific case of large granular lymphocytic leukemia, the highest incidence in the treated animals was similar to that in untreated controls.

Dose descriptor:

NOEL

Effect level:

> 2.6 other: mg/cm² (> 15 % BPO)

Based on:

act. ingr.

Sex:

male/female

Basis for effect level:

other: No carcinogenic effects were observed.

Remarks on result:

other: Effect type: carcinogenicity (migrated information)

Dose descriptor:

LOAEL

Remarks:

local effects

Effect level:

0.3 other: mg/cm² (1.67% BPO)

Based on:

act. ingr.

Sex:

male/female

Basis for effect level:

other: The major microscopic findings were observed at the application site from 1.67%.: hyperkeratosis, subepidermal subacute inflammation , and sebaceous gland hyperplasia

Remarks on result:

other: Effect type: toxicity (migrated information)

Dose descriptor:

NOAEL

Remarks:

systemic effects

Effect level:

> 2.6 other: mg/cm² (> 15 % BPO)

Based on:

act. ingr.

Sex:

male/female

Basis for effect level:

other: No systemic toxicity was observed

Remarks on result:

other: Effect type: toxicity (migrated information)

Conclusions:

In conclusion, under the conditions of the study, there were no findings indicative of oncogenicity resulting from daily topical exposure of rats to benzoyl peroxide gels at concentrations up to 15% (dose level of 45 mg/rat/day) for 104 consecutive weeks.

Executive summary:

The oncogenic potential of benzoyl peroxide gels was evaluated when administered daily by topical application to the dorsal skin of rats for at least 104 weeks. Male and female CDF®(F-344)Cr1BR rats were assigned to seven groups. Benzoyl peroxide in carbopol gel at concentrations of 1.67, 5.0, and 15.0% (5, 15. and 45 mg/rat/day) was applied topically once daily to a treatment area (approximately 3.5 x 5 cm) on the dorsal skin of rats in Groups 2, 3, and 4, respectively. Rats in Group 1 served as vehicle controls and received daily topical applications of the carbopol gel vehicle at a dose volume of 0.3 mL. Rats in Group 6 served as negative controls; the hair on the backs of these rats was clipped at the same intervals as the other rats on study; however, these rats were not treated. Sixty rats/sex were assigned to Groups 1, 2, 3, 4, and 6, with the first 10 rats/sex/group designated for interim sacrifice during Week 53 and the remaining 50 rats/sex/group designated for terminal sacrifice after 104 weeks of treatment. Fifty rats/sex in Group 5 served as recovery animals, in that they were treated with 15% benzoyl peroxide for 52 weeks, then treated with the vehicle for the remainder of the study. Twenty rats/sex in Group 7 served as sentinel animals for pathogen screening at Weeks 26, 52, 78, and 104. Diet and water were providedad libitum. Once weekly, each animal was removed from its cage and examined for abnormalities and signs of toxicity, specifically noting the location, size, and appearance of any grossly visible or palpable masses. The treated skin (or analogous site on the untreated control) was graded for irritation once weekly. Body weights were recorded weekly from Weeks 1 through 14 and every fourth week thereafter and at Weeks 53 and 105. Food consumption was measured and recorded weekly for Weeks 1 through 13 and every fourth week thereafter and at Weeks 52 and 104. Blood smears were prepared from all moribund-, interim-, and terminal-sacrifice animals for possible evaluation of hematopoietic neoplasia. After 52 weeks (interim-sacrifice animals) or 104 weeks (terminal sacrifice animals) of treatment, the animals were anesthetized, weighed, exsanguinated, and necropsied. At necropsy, macroscopic observations were recorded, and selected tissues were collected and preserved. The liver, kidneys, and brain were weighed from all animals at interim sacrifice. Selected tissues (treated skin, untreated skin, and kidneys) were examined microscopically from all interim-sacrifice rats (Groups 1, 2, 3. 4. and 6). All tissues were examined microscopically from all terminal-sacrifice rats in Groups 1, 2. 3. 4, and 6, whereas only treated and untreated skin was examined microscopically from the rats in Group 5. Tumors were statistically analyzed separately and combined for relationship to treatment.

Treatment had no effect on: survival, body weights, food consumption, clinical observations, application site findings, gross pathology findings and organ weight changes at the interim sacrifice and gross pathology findings at terminal sacrifice. Microscopic evaluation revealed that the only treatment-related histopathologic findings noted at both the interim and terminal sacrifices were those seen at the site of topical dermal exposure. These consisted of hyperkeratosis, acanthosis, sebaceous gland hyperplasia, and chronic subepidermal inflammation. These changes occurred in all treatment groups (1.67, 5, and 15% benzoyl peroxide) and were mostly of a minimal to mild degree of severity. They progressed only slightly between 52 and 104 weeks of exposure. For the recovery animals (52 weeks of treatment with 15% benzoyl peroxide followed by 52 weeks of vehicle), hyperkeratosis was present approximately twice as frequently as it was in either the vehicle or negative control group, but less than half as frequently as in any other treatment group. Acanthosis was not appreciably increased in the recovery animals as compared to controls, and sebaceous gland hyperplasia was only equivocally present. These findings indicated good, although not complete, recovery from the effects of 52 weeks of treatment with 15% benzoyl peroxide, with remnants of hyperkeratosis remaining after 52 weeks of recovery.

In conclusion, under the conditions of the study, there were no findings indicative of oncogenicity resulting from daily topical exposure of rats to benzoyl peroxide gels at concentrations up to 15% (dose level of 45 mg/rat/day, 2.6 mg/cm²) for 104 consecutive weeks.

Endpoint conclusion

Endpoint conclusion:

no adverse effect observed

Study duration:

chronic

Species:

other: rat, mouse

Justification for classification or non-classification

According to EU directive 67/548/EEC and EU Regulation (EC) N0. 1272/2008 (CLP), the substance is not classified for carcinogenicity.

Additional information

Complete carcinogenicity studies

Oral administration

Mouse: Groups of 25 male and 25 female albino mice (strain and age unspecified] were fed a diet containing different doses of a commercial powder containing 18% benzoyl peroxide [purity unspecified] which were based upon consideration of the use of benzoyl peroxide in bread production (calculated doses of benzoyl peroxide: 0, 28, 280 and 2800 mg/kg of diet), for 80 weeks, at which time the number of survivors was 3, 10, 0 and 2 male mice and 9, 11, 9 and 11 females; all surviving animals were then killed. A few tumors at various sites were observed, but the overall tumor incidence did not differ significantly between treated and control groups (Sharratt et al., 1964).

Rat: Groups of 25 male and 25 female albino rats [strain and age unspecified] were fed a diet containing different doses of a commercial powder containing 18% benzoyl peroxide [purity unspecified] which were based upon consideration of the use of benzoyl peroxide in bread production (calculated doses of benzoyl peroxide: 0, 28, 280 and 2800 mg/kg of diet) for 120 weeks, at which time they were killed. At 104 weeks, 12, 12, 13 and 11 males and 14, 7, 9 and 11 females survived. A few tumors were observed at various sites, but the over­ all tumor incidence did not differ significantly between treated and control groups (Sharratt et al., 1964).

Dermal route

Two-year dermal carcinogenicity studies conducted according to GLP found no evidence of oncongenicity resulting from daily topical application of a benzoyl peroxide gel in rats and mice at doses that meet the maximum tolerance dose.

Rat: The oncogenic potential of benzoyl peroxide gels was evaluated when administered daily by topical application to the dorsal skin of rats for at least 104 weeks (CHPA, 2002; Minnema, 2000). Male and female CDF®(F-344) Cr1BR rats were assigned to seven groups. Benzoyl peroxide in carbopol gel at concentrations of 1.67, 5.0, and 15.0% (5, 15. and 45 mg/rat/day) was applied topically once daily to a treatment area (approximately 3.5 x 5 cm) on the dorsal skin of rats in Groups 2, 3, and 4, respectively. Rats in Group 1 served as vehicle controls and received daily topical applications of the carbopol gel vehicle at a dose volume of 0.3 mL. Rats in Group 6 served as negative controls; the hair on the backs of these rats was clipped at the same intervals as the other rats on study; however, these rats were not treated. Sixty rats/sex were assigned to Groups 1, 2, 3, 4, and 6, with the first 10 rats/sex/group designated for interim sacrifice during Week 53 and the remaining 50 rats/sex/group designated for terminal sacrifice after 104 weeks of treatment. Fifty rats/sex in Group 5 served as recovery animals, in that they were treated with 15% benzoyl peroxide for 52 weeks, then treated with the vehicle for the remainder of the study. Twenty rats/sex in Group 7 served as sentinel animals for pathogen screening at Weeks 26, 52, 78, and 104. Diet and water were provided ad libitum. Once weekly, each animal was removed from its cage and examined for abnormalities and signs of toxicity, specifically noting the location, size, and appearance of any grossly visible or palpable masses. The treated skin (or analogous site on the untreated control) was graded for irritation once weekly. Body weights were recorded weekly from Weeks 1 through 14 and every fourth week thereafter and at Weeks 53 and 105. Food consumption was measured and recorded weekly for Weeks 1 through 13 and every fourth week thereafter and at Weeks 52 and 104. Blood smears were prepared from all moribund-, interim-, and terminal-sacrifice animals for possible evaluation of hematopoietic neoplasia. After 52 weeks (interim-sacrifice animals) or 104 weeks (terminal sacrifice animals) of treatment, the animals were anesthetized, weighed, exsanguinated, and necropsied. At necropsy, macroscopic observations were recorded, and selected tissues were collected and preserved. The liver, kidneys, and brain were weighed from all animals at interim sacrifice. Selected tissues (treated skin, untreated skin, and kidneys) were examined microscopically from all interim-sacrifice rats (Groups 1, 2, 3. 4. and 6). All tissues were examined microscopically from all terminal-sacrifice rats in Groups 1, 2. 3. 4, and 6, whereas only treated and untreated skin was examined microscopically from the rats in Group 5. Tumors were statistically analyzed separately and combined for relationship to treatment.

Treatment had no effect on: survival, body weights, food consumption, clinical observations, application site findings, gross pathology findings and organ weight changes at the interim sacrifice and gross pathology findings at terminal sacrifice. Microscopic evaluation revealed that the only treatment-related histopathologic findings noted at both the interim and terminal sacrifices were those seen at the site of topical dermal exposure. These consisted of hyperkeratosis, acanthosis, sebaceous gland hyperplasia, and chronic subepidermal inflammation. These changes occurred in all treatment groups (1.67, 5, and 15% benzoyl peroxide) and were mostly of a minimal to mild degree of severity. They progressed only slightly between 52 and 104 weeks of exposure. For the recovery animals (52 weeks of treatment with 15% benzoyl peroxide followed by 52 weeks of vehicle), hyperkeratosis was present approximately twice as frequently as it was in either the vehicle or negative control group, but less than half as frequently as in any other treatment group. Acanthosis was not appreciably increased in the recovery animals as compared to controls, and sebaceous gland hyperplasia was only equivocally present. These findings indicated good, although not complete, recovery from the effects of 52 weeks of treatment with 15% benzoyl peroxide, with remnants of hyperkeratosis remaining after 52 weeks of recovery.

In conclusion, under the conditions of the study, there were no findings indicative of oncogenicity resulting from daily topical exposure of rats to benzoyl peroxide gels at concentrations up to 15% (dose level of 45 mg/rat/day, 2.6 mg/cm²) for 104 consecutive weeks.

Mouse: The oncogenic potential of benzoyl peroxide gels was evaluated when administered daily by topical application to the dorsal skin of mice for at least 104 weeks (CHPA 2001; Minnema, 2001). Male and female B6C3F1/Cr1BR mice were assigned to seven groups. Benzoyl peroxide in carbopol gel at concentrations of l, 5, and 25-15% (1, 5, and 25-15 mg/mouse/day) was applied topically once daily to a treatment area (approximately 2 x 3 cm) on the dorsal skin of mice in Groups 2, 3, and 4, respectively. Mice in Group 1 served as vehicle controls and received daily topical applications of the carbopol gel vehicle at a dose volume of 0.1 mL. Mice in Group 6 served as negative controls; the hair on the backs of these mice was dipped at the same intervals as the other mice on study; however, these mice were not treated. Sixty mice/sex were assigned to Groups 1, 2, 3, 4, and 6, with the first 10 mice/sex/group designated for interim sacrifice during Week 53 and the remaining 50 mice/sex/group designated for terminal sacrifice after 104 weeks of treatment. Fifty mice/sex in Group 5 served as recovery animals, in that they were treated with 25% benzoyl peroxide for 52 weeks, and then treated with the vehicle for the remainder of the study. Twenty mice/sex in Group 7 served as sentinel animals for pathogen screening at Weeks 26, 52, 78, and 104. Diet and water were provided ad libitum. Once weekly, each animal was removed from its cage and examined for abnormalities and signs of toxicity, specifically noting the location, size, and appearance of any grossly visible or palpable masses. The treated skin (or analogous site on the untreated control) was graded for irritation once weekly. Body weights were recorded weekly from Weeks l through 14 and every fourth week thereafter and at Weeks 53 and 105. Group 4 had additional body weights taken at Weeks 57 and 59. Food consumption was measured and recorded weekly for Weeks l through 13 and every fourth week thereafter and at Weeks 52 and 104. Blood smears were prepared from all moribund-, interim-, and terminal-sacrifice animals for possible evaluation of hematopoietic neoplasia. After 52 weeks (interim-sacrifice animals) or 104 weeks (terminal sacrifice animals) of treatment the animals were anesthetized, weighed, exsanguinated, and necropsied. At necropsy, macroscopic observations were recorded, and selected tissues were collected and preserved. The liver with gallbladder, kidneys, and brain were weighed from all animals at interim sacrifice. Selected tissues (treated skin, untreated skin, and livers) were examined microscopically from all interim-sacrifice mice (Groups 1, 2, 3, 4, and 6). All collected tissues were examined microscopically at all terminal-sacrifice mice in Groups 1, 2, 3, 4, and 6, whereas only treated and untreated skin was examined microscopically from the mice in Group 5 (high dose-discontinued). Tumors were statistically analyzed separately and combined for relationship.

Daily topical exposure of mice to benzoyl peroxide at concentrations of 1 and 5% continued for 104 weeks. However, due to findings (skin ulcerations at the application site) that 25% benzoyl peroxide exceeded the maximum tolerated dose, the concentration in the high-dose group was lowered to 15% at the beginning of Week 57. Due to further incidences of skin ulceration at the 15% benzoyl peroxide concentration, treatment of the high-dose animals was discontinued (with vehicle only) for the final 13 weeks of the study. With the exception of the findings of ulcerations at the application site, there were no treatment-related differences in clinical observations among any of the groups. Treatment did not affect survival, body weights, or food consumption. The major microscopic findings were observed at the application site. At the interim (Week 53) sacrifice, these findings consisted of hyperkeratosis (1, 5, and 25% benzoyl peroxide), subepidermal subacute inflammation (5 and 25% benzoyl peroxide), and sebaceous gland hyperplasia (males: 5 and 25% benzoyl peroxide: females:1, 5, and 25% benzoyl peroxide). These findings were dose-dependent with regards to incidence and or group mean severity. Similar findings were noted at the terminal (Week 105-106) sacrifice with the exception that there were no findings for treated or untreated skin in the high-dose and high-dose-discontinued animals. The high-dose-discontinued animals, after being treated with 25% benzoyl peroxide for 1 year, were allowed 52 weeks of recovery. Although the high-dose animals were not originally intended to be a recovery group, they were net treated with benzoyl peroxide for the final 13 weeks of the study since the 25 -15% benzoyl peroxide concentrations had been found to exceed the maximum tolerated dose. In both of these cases there was no residual effect of treatment.

In conclusion, under the conditions of this study, there were no histologic findings indicative of oncogenicity resulting from daily topical exposure of mice to benzoyl peroxide gels at concentrations up to 25% (dose level 25 mg/mouse /day; 4.17 mg/cm²). The spectrum of neoplasms observed in this study was typical for aging B6C3F1 mice.

Additional carcinogenicity data

Additional carcinogenicity data available on benzoyl peroxide were reviewed by IARC (1985, 1999) and the Health Council of the Netherlands (2012). Benzoyl peroxide was tested for carcinogenicity in mice and rats by subcutaneous administration, and in mice by skin application. In studies by skin application in mice, benzoyl peroxide was tested for either initiating or promoting activity. All of the studies were inadequate for an evaluation of complete carcinogenicity; two studies indicated that benzoyl peroxide has promoting activity in mouse skin.

Subcutaneous and/or intramuscular administration

Mouse: Groups of 25 male and 25 female albino mice (age and strain unspecified) received a single subcutaneous injection of 50 mg of a 20% suspension of benzoyl peroxide (purity unspecified] in starch solution and were observed for 80 weeks, at which time they were killed. An equal number of animals was injected with starch solution only and served as vehicle controls. Nine treated males, seven treated females, no control male and six control females survived. No tumor was found at the site of injection or at any other site in treated or control groups (Sharratt et al., 1964). [The IARC Working Group noted the short duration of the experiment, the high mortality and the low degree of exposure.]

Rat: A group of 25 male and 25 female albino rats (age and strain unspecified) received a single subcutaneous injection of 120 mg of a 20% suspension of benzoyl peroxide [purity unspecified] in starch solution and were observed for 120 weeks, at Which time they were killed. An equal number of animals was injected with starch solution only and served as vehicle controls. Ten treated males and 16 control males and nine treated females and 17 control females survived 104 weeks. No tumor was found at the site of injection, and overall tumor incidence did not differ significantly between treated and control groups (Sharratt et al., 1964). [The IARC Working Group noted the low degree of exposure.]

A group of 20 male Charles River CD rats [age unspecified] received subcutaneous injections of 2.9 mg benzoyl peroxide [purity unspecified] in 0.2 ml trioctanoin (tricaprylin) into the right hind leg twice weekly for 12 weeks and were observed for 14 months, at which time all animals were still alive. A further group of 20 male rats received trioctanoin only and served as vehicle controls. No malignant tumor was found at the injection site or in internal organs (Poirier et al., 1967). [The IARC Working Group noted the short duration of the experiment.]

Groups of 20 male and 15 female Bethesda black rats (NIH Black rats) received a subcutaneous implantation of 50 mg benzoyl peroxide [purity unspecified] in a gelatin capsule at the nape of the neck and were observed for 24 months. Mortality was 9/35 (sex unspecified] at week 52 and 22/35 at week 78. None of the rats developed tumors at the site of implantation. Single tumors were found at various other sites, but none of the tumors was considered by the author to be causally related to benzoyl peroxide treatment (Hueper, 1964).

Skin application

Mouse: A group of 30 male ICR/Ha mice, eight weeks old, received thrice-weekly applications for life of approximately 100 mg benzoyl peroxide [purity unspecified] dissolved in 5% benzene. Median survival time was 292 days; one mouse developed a skin papilloma. A total of 150 mice (from four different control groups) were treated with benzene alone. Median survival times ranged from 262-412 days; 11 skin tumors, including one carcinoma, were observed (Van Duuren et al., 1963).

Groups of 25 male and 25 female albino mice [age and strain unspecified] received skin applications of one drop (approximately 50 mg) of a 50% suspension of benzoyl peroxide [purity unspecified] in flour paste on the back of the neck on six days per week for 80 weeks. An equal number of mice painted with flour paste only served as controls. No skin tumor was observed, and overall tumor incidence did not differ significantly between treated groups and controls (Sharratt et al., 1964). [The IARC Working Group noted that adequate data on survival were not given.]

A group of 21 female Swiss mice [age unspecified] received twice-weekly skin applications of 0.5% benzoyl peroxide [purity unspecified] in acetone for 80 weeks. No skin tumor was observed. Another group of 20 females received applications of 0.5% benzoyl peroxide in acetone twice weekly for three weeks then, after one week, they were treated with 5% croton oil in mineral oil twice weekly for 67 weeks. No skin tumor was reported (Saffiotti & Shubik, 1963). [The IARC Working Group noted that adequate data on survival were not given.]

Mouse: A group of 20 female SEN mice, four weeks of age, was treated twice weekly for 51 weeks with 0.2 mL of a 100 mg/mL solution of benzoyl peroxide in acetone applied to the skin shaved 48 h previously. A group of 15 mice receiving 0.2 mL acetone served as controls. At the termination of the experiment, there were no skin tumors among the control mice, compared with 8/20 in the benzoyl peroxide-treated mice (p < 0.05), of which 5/20 were squamous-cell carcinomas. The first tumor developed in week 24. Six of 20 mice showed epidermal hyperplasia (Kurokawa et al., 1984). According to Kraus et al. (1995), differences between the study by Kurokawa et al. (1984) and the studies conducted by other workers include the use of different bedding, daily shaving of the application site of the mice by Kurokawa et al. (1984), as opposed to shaving only at the start of the study, and possible differences in criteria used by study pathologists for assessment of tumors (Slaga, 1992). It is not known to what extent these differences may have contributed to the results reported by Kurokawa, compared to those re­ ported by others (Slaga, 1992). The multiplicity of papillomas and the incidence of carcinomas reported by Kurokawa et al. (1984), using the standard DMBA initiation, TPA promotion protocol, were extremely high relative to findings by other workers using this protocol (Athar et al., 1990; NTP, 1994; Slaga, 1982 and 1984, Slaga et al., 1981). This may indicate that the SENCAR mice used by Kurokawa et al. (1984) had an unusually high background of initiated cells. Because a control group treated only with the skin tumor promoter TPA was not included in this study, it is impossible to establish the basis of the high response observed in the DMBA-TPA-treated group and hence the possibility that benzoyl peroxide promoted initiated cells in this study. Thus, the findings of Kurokawa et al. (1984) must be considered uninterpretable in light of the lack of a TPA-only control and the disparity with the results of other workers using the standard DMBA­TPA protocol. Clearly the results of Kurokawa et al. (1984) are at odds with the overwhelming negative evidence from other studies.

Groups of five male heterozygous TG. AC mice (carrying a v-Ha-ras gene) derived from the wild-type FVB/N strain were treated with 0, 1, 5 or 10 mg benzoyl peroxide in 0.2 mL acetone on the shaved dorsal skin twice a week for 20 weeks. Groups of five male FVB/N mice were similarly treated. No papillomas developed in the FVB/N mice. The incidences of papilloma-bearing mice in the four groups of TG: AC mice were 0/5, 0/5, 3/5 and 3/4, respectively (one papilloma-bearing mouse in the 10-mg group died before the end of the experiment) (Spalding et al., 1993).

Initiation/Promotion studies

Groups of 30 female Sencar mice, aged seven to nine weeks, were used to test the tumor-promoting (A), tumor-initiating (B) and complete carcinogenic (C) activities of benzoyl peroxide [purity unspecified] on the skin. Mice in experiment (A) received a single topical application of 10 nmol 7,12-dimethylbenz[a]anthracene (DM8A) in 0.2 ml acetone, followed one week later by applications of 1, 10, 20 or 40 mg benzoyl peroxide in 0.2 ml acetone twice weekly for 52 weeks. A group receiving a single application of 0.2 ml acetone alone served as controls. The numbers of mice with papillomas at 30 weeks were 1/28 {controls), 9/29 (1 mg), 20/28 (10 mg), 21/27 (20 mg) and 20/24 (40 mg). The numbers of mice with carcinomas at 52 weeks in these groups were 0/28, 1/29, 6/28, 12/27 and 10/24, respectively. Mice in experiment (8) received a single topical application of 0.2 ml acetone alone or 1, 10, 20 or 40 mg benzoyl peroxide in acetone, followed one week later by twice-weekly applications of 2 µg 12-0-tetradecanoylphorbol 13-acetate (TPA) in acetone for 52 weeks. No significant difference was observed in the incidence of papillomas; no carcinoma was observed. Mice in experiment {C) received twice-weekly topical applications of acetone alone or 1, 10, 20 or 40 mg benzoyl peroxide in 0.2 ml acetone for 52 weeks. No significant difference was observed in the incidence of papillomas; no carcinoma was observed. Survival rates in all experiments at 30 weeks ranged from 24-29 mice per group (Slaga et al., 1981). [The IARC Working Group noted the absence of a control group treated with DM8A only, and the short duration of the experiment for complete carcinogenicity.]

Groups of 30-40 female Sencar mice, aged seven to eight weeks, received a single application of 10 nmol DM8A in 0.2 ml acetone and, one week later, were treated twice weekly either with 2 µg TPA in 0.2 ml acetone or 20 mg benzoyl peroxide [purity unspecified] in 0.2 ml acetone for 48 weeks. The incidence of skin papillomas was 100% at 20 weeks in the TPA group and about 80% at 48 weeks in the benzoyl peroxide group. The incidence of squamous-cell carcinomas was 0% in the TPA group and about 50% in the benzoyl peroxide group. Additional groups of 30-40 female C57BL/6 mice, aged seven to eight weeks, received a single skin application of 400 or 800 nmol DM8A or 1600 or 3200 nmol benzo[a]pyrene {BP) in 0.2 ml acetone, and one week later were treated twice weekly with 4 µg TPA in 0.2 ml acetone or 20 mg benzoyl peroxide in 0.2 ml acetone for 48 weeks. The incidence of squamous-cell carcinomas was about 60% in the DMBA-benzoyl peroxide group, <50% in the DMBA-TPA group, >50% in the BP-benzoyl peroxide group and <50% in the BP-TPA group (Reiners et al., 1984). [Those values were taken from graphs, and actual numbers were not given. The IARC Working Group noted that survival rates were not given, nor were data available on controls treated with DM8A or BP.]

Mouse: Benzoyl peroxide was tested for promoting activity in groups of 20 and 15 female SEN mice receiving a single topical application of 20 nmol 7,12-dimethylbenz[a]anthracene (DMBA) followed by either 0.2 mL of a 100 mg/mL solution of benzoyl peroxide in acetone or acetone alone for 51 weeks. At the termination of the experiment, there were no skin tumors among the 15 control mice, compared with 20/20 in the benzoyl peroxide-treated mice (p < 0.01), of which 18/20 were squamous-cell carcinomas. The first tumor developed in week 8. All 20 treated mice showed epidermal hyperplasia (Kurokawa et al., 1984). [Differences between the study by Kurokawa et al. (1984) and the studies conducted by other workers include the use of different bedding, daily shaving of the application site of the mice by Kurokawa et al. (1984), as opposed to shaving only at the start of the study, and possible differences in criteria used by study pathologists for assessment of tumors (Slaga, 1992). It is not known to what extent these differences may have contributed to the results reported by Kurokawa, compared to those reported by others (Slaga, 1992). The multiplicity of papillomas and the incidence of carcinomas reported by Kurokawa et al. (1984), using the standard DMBA initiation, TPA promotion protocol, were extremely high relative to findings by other workers using this protocol (Athar et al., 1990; NTP, 1994; Slaga, 1982 and 1984, Slaga et al., 1981). This may indicate that the SENCAR mice used by Kurokawa et al. (1984) had an unusually high background of initiated cells. Because a control group treated only with the skin tumor promoter TPA was not included in this study, it is impossible to establish the basis of the high response observed in the DMBA-TPA-treated group and hence the possibility that benzoyl peroxide promoted initiated cells in this study. Thus, the findings of Kurokawa et al. (1984) must be considered uninterpretable in light of the lack of a TPA-only control and the disparity with the results of other workers using the standard DMBA­TPA protocol. Clearly the results of Kurokawa et al. (1984) are at odds with the overwhelming negative evidence from other studies (Kraus et al., 1995).

Groups of female SEN mice, five to seven weeks of age, were treated with a single topical application of 10 nmol DMBA on the shaved dorsal skin. Twice-weekly applications of 1µg 12-O-tetradecanoylphorbol 13-acetate (TPA) were begun two weeks later and continued for 20 weeks. Beginning at week 21, one group of 21 papilloma-bearing mice continued to receive 1µg TPA, while another group of 20 papilloma-bearing mice began twice-weekly treatments of 20 mg benzoyl peroxide. All solutions were applied in 0.2 mL acetone and treatments were ended at week 40. No new tumors appeared during weeks 21–40 in the benzoyl peroxide-treated group. At the end of the experiment, the proportion of mice with skin carcinomas was 70% in the benzoyl peroxide-treated group compared with 38% in the TPA-treated group and the cumulative number of carcinomas was 3.25-fold higher in the benzoyl peroxide-treated group. All skin tumors present at the end of the experiment were examined histologically. While no keratoacanthomas were identified in the TPA-treated group, 17 were found in the benzoyl peroxide-treated group. The authors concluded that benzoyl peroxide enhances the progression of benign to malignant tumors (O’Connell et al., 1986).

However, in a similarly designed and executed experiment, it was found that benzoyl peroxide did not enhance the progression of papillomas to squamous-cell carcinomas in SEN mice (Battalora et al., 1996). In this study, athree-stage treatment protocol was utilized to test the ability of BPO to enhance the progression of pre-existing papillomas to squamous cell carcinomas (SCCs). In stage I, mice were initiated with 0.5 (µg of DMBA. In stage II, mice were promoted with twice-weekly applications of 1 or 2 (µg of 12-0-tetradecanoylphorbol-13-acetate (TPA) for 15 weeks. Then, in stage III, mice were treated with acetone, TPA (1 or 2 µg) or benzoyl peroxide (BPO; 20 mg) for the next 45 weeks. The mean number of papillomas per mouse at plateau was very similar for all groups. The carcinoma incidence was also similar for all groups regardless of the treatment protocol used, as was the mean number of carcinomas per mouse. BPO did not enhance the progression of papillomas to SCCs under the current experimental conditions.

Three groups of 16 male and 16 female hr/hr Oslo strain mice [age unspecified] were treated with a single topical application of 51.2µg DMBA in 100µL acetone. One group then received an application (rubbed into the skin [quantity not specified]) of Panoxyl, a gel containing 5% benzoyl peroxide used for the treatment of acne, twice each week for up to 60 or 61 weeks; a second group was similarly treated, but with gel not containing benzoyl peroxide and the third group was not treated further. A fourth group received Panoxyl treatment only. The total numbers of skin tumors in each group of mice were: DMBA alone, 22/32 (18 papillomas, 4 squamous-cell carcinomas); Panoxyl alone, 2/32 (2 squamous-cell carcinomas); DMBA + Panoxyl, 51/30 (49 papillomas, 3 squamous- cell carcinomas); DMBA + gel, 31/32 (31 papillomas). Both Panoxyl and the gel without benzoyl peroxide increased the multiplicity of papillomas induced by DMBA (Iversen, 1986).

The last experiment was repeated in part using hr/hr Oslo strain mice and extended by the use of SENCAR mice. The data were mainly presented as summary statistics. In contrast to the earlier results, the gel without benzoyl peroxide did not enhance DMBA carcinogenesis in hr/hr mice and Panoxyl did not enhance DMBA carcinogenesis in SEN mice. Groups of 32 hr/hr mice were also treated with ultraviolet radiation (UV) from new Phillips HP 3114 sunlamps, either alone twice a week or 5–30 min before treatment with Panoxyl or the gel without benzoyl peroxide. The numbers of mice with skin carcinomas (and the total numbers of skin tumors in each group of mice) were: UV alone, 26/32 (103 papillomas, 45 carcinomas); UV + Panoxyl, 22/32 (94 papillomas, 30 carcinomas); UV + gel, 23/32 (126 papillomas, 29 carcinomas). Thus, neither Panoxyl nor the gel without benzoyl peroxide had any significant effect upon the multiplicity of UV-induced skin tumors (Iversen, 1988).

One hundred and forty-eight Uscd (Hr) albino hairless mice [sex unspecified], three to four months of age, received 270 mJ/cm2 UVB radiation to the posterior halves of their backs three times each week for eight weeks. The UVB source was an Hanovia air-cooled hot quartz contact lamp emitting 54 mJ/cm²/s of UVB energy at a distance of 3.4 cm. Four weeks later the irradiated mice were divided into four groups that were treated in the irradiated area as follows: group 1 received 0.1 mL 0.1% croton oil in acetone five times per week for the duration of the study; group 2 received 0.1 mL acetone; group 3 received 0.1 mL benzoyl peroxide diluent; group 4 received 0.1 mL 5% benzoyl peroxide lotion [i. e., about 5 mg benzoyl peroxide]. At week 62, when the experiment was terminated, the tumor incidences in the irradiated areas of the skin were: group 1, 9/24; group 2, 1/20; group 3, 1/22; group 4, 2/26. Under the circumstances of the experiment, croton oil, but not benzoyl peroxide, enhanced the incidence of skin tumors induced by UVB radiation (Epstein, 1988).

Iron overload augments benzoyl peroxide (BPO) -mediated tumor promotion in 7,12-dimethylbenz[a]anthracene (DMBA) -initiated mouse skin. Female albino Swiss mice were overloaded with iron and tumors were initiated by applying a single topical application of DMBA. A week after the initiation, promoting agent, BPO, was applied three times/week for 46 weeks. The appearance of the first tumor (papilloma) and the number of tumors/mouse were recorded. When compared to the control group, the iron-overloaded mice showed an increased incidence of tumors at various time intervals. In iron-overloaded animals, tumors appeared earlier and also the number of tumors/mouse was significantly higher. These data could be correlated with the iron levels of mouse skin in the two groups. Further, BPO-mediated induction in ornithine decarboxylase (ODC) activity and [3H]thymidine incorporation in cutaneous DNA were higher in the iron overload group. In addition, in iron-overloaded mice, cutaneous lipid peroxidation (LPO) and xanthine oxidase (XOD) activities were higher, whereas catalase activity was reduced. Similar to papilloma induction, a significant increase in carcinoma yield and incidence was observed in iron-overloaded animals. Based on this study, it appears that iron overload significantly increases the tumor promotion and progression potential of BPO. Oxidative stress generated by iron overload could be responsible for the augmentation of BPO-mediated cutaneous tumorigenesis (Rezazadeh and Athar, 1988).

A comparative initiation/promotion skin application study was conducted with B6C3F1, CD-1 and SEN mice (NTP, 1996). In the portions of the study that are relevant to benzoyl peroxide, groups of 30 male and 30 female mice of each strain were treated on the shaved skin as follows: A, acetone, the vehicle for all the substances (0.1 mL) alone, once per week; B, 2.5µg DMBA, once, followed one week later with 0.1 mL acetone once per week for 51 weeks; C, 25µg DMBA, once, followed one week later with 0.1 mL acetone once per week for 51 weeks; D, 20 mg benzoyl peroxide, followed one week later with 20 mg benzoyl peroxide in 0.2 mL acetone once per week for 51 weeks; E, as group B, followed one week later with 20 mg benzoyl peroxide in 0.2 mL acetone once per week for 51 weeks; F, as group C, followed one week later with 20 mg benzoyl peroxide in 0.2 mL acetone once per week for 51 weeks; G, 100µg N-methyl-N’-nitro-N-nitrosoguanidine (MNNG), once, followed one week later with 0.1 mL acetone once per week for 51 weeks; H, 500µg MNNG, once, followed one week later with 0.1 mL acetone once per week for 51 weeks; I, as group D; J, as group G, followed one week later with 20 mg benzoyl peroxide in 0.2 mL acetone once per week for 51 weeks; and K, as group H, followed one week later with 20 mg benzoyl peroxide in 0.2 mL acetone once per week for 51 weeks. For each strain, survival of male and female mice in most groups was similar, but it was significantly reduced (p < 0.01) in male CD-1 and SEN mice of group H. Body weight gain also was similar in most groups, but was significantly reduced in female B6C3F1 mice of group K (p < 0.01). The skin tumor responses are shown in the following. In neither sex of any strain did benzoyl peroxide act as a complete skin carcinogen, but it was active as a promoter in both sexes of SEN mice, following initiation with DMBA (at both dose levels) or MNNG (at both dose levels). The CD-1 strain and, in particular, the B6C3F1 strain, were clearly less sensitive than the SEN strain.

Table. Skin tumour responses to treatment with carcinogens and/or benzyl peroxide in mice

Sex, group (treatment)	B6C3F1	CD-1	SEN
Male A (acetone)	0/30	0/30	0/30
Male B (DMBA 2.5 µg)	0/30	0/30	0/29
Male C (DMBA 25 µg)	0/30	0/30	0/31
Male D (benzoyl peroxide/benzoyl peroxide)	0/30	0/30	0/30
Male E (DMBA 2.5 µg/benzoyl peroxide)	1/30	1/30	20/30
Male F (DMBA 25 µg/benzoyl peroxide)	1/30	6/30	22/30
Male G (MNNG 100 µg)	0/30	1/30	2/30
Male H (MNNG 500 µg)	1/30	7/30	19/30
Male I (benzoyl peroxide /benzoyl peroxide)	0/30	0/30	0/30
Male J (MNNG 100 µg/benzoyl peroxide)	0/30	1/30	9/30
Male K (MNNG 500 µg/benzoyl peroxide)	3/30	11/30	25/30
Female A (acetone)	0/30	0/30	0/29
Female B (DMBA 2.5 µg)	0/30	0/30	0/31
Female C (DMBA 25 µg)	0/30	0/30	0/29
Female D (benzoyl peroxide/benzoyl peroxide)	0/30	0/30	0/30
Female E (DMBA 2.5 µg/benzoyl peroxide)	4/30	1/30	22/30
Female F (DMBA 25 µg/benzoyl peroxide)	2/30	5/30	20/30
Female G (MNNG 100 µg)	0/30	0/30	0/30
Female H (MNNG 500 µg)	0/30	6/30	8/30
Female I (benzoyl peroxide/benzoyl peroxide)	0/30	0/30	1/30
Female J (MNNG 100 µg/benzoyl peroxide)	1/30	3/30	9/30
Female K (MNNG 500 µg/benzoyl peroxide)	3/30	13/30	16/30

From NTP (1996)

DMBA, 7,12-dimethylbenz[a]anthracene; MNNG, N-methyl-N'-nitro-N-nitrosoguanidine of benzoyl peroxide were 20 mg.

Hamster: Male Syrian hamsters [age unspecified] were randomized into five groups of 20 and treated as follows: group 1 received 1 mL acetone applied to the shaved dorsal area three times each week; group 2 received 10 mg/kg bw DMBA in sesame oil once by gavage; group 3 received 160 mg benzoyl peroxide in 1 mL acetone applied to the shaved dorsal skin three times per week; group 4 was treated with DMBA as group 2, followed one week later by 80 mg benzoyl peroxide in 1 mL acetone applied to the shaved dorsal skin three times per week; group 5 was treated with DMBA as group 2, followed one week later by treatment as group 3. After 16 months, all surviving hamsters were killed. The 25% survival in the groups was: group 1, 442 days; group 2, 376 days; group 3, 427 days; group 4, 342 days; group 5, 407 days. Histological assessment of the treated areas of skin indicated that the numbers of melanotic foci per hamster (arithmetic means and 95% CI) were: group 2, 6.3 (3.6–9.0); group 4, 17.4 (13.2–21.6); group 5, 26.9 (22.5–31.2); the corresponding numbers of melanotic tumors per hamster were: group 2, 0.6 (0.2–1.0); group 4, 2.2 (1.3–3.1); group 5, 2.9 (2.0–3.7). Benzoyl peroxide treatment enhanced the frequency of melanotic skin tumors in Syrian hamsters treated with DMBA (Schweizer al., 1987).

A subthreshold dose of 0.1% DMBA in oil applied to hamster buccal pouch three times a week for 10 weeks did not result in the development of carcinoma in up to 25 weeks from the start of the experiment. However, if the subthreshold dose of DMBA was followed by 6 weeks of no treatment and six weeks of 40% benzoyl peroxide treatment, carcinomas developed rapidly. Both benzoyl peroxide and its acetone solvent are noncarcinogenic. An experiment carried out on sixty-six young adult male and female Syrian hamsters demonstrated these results, using appropriate controls (Odukaya and Shklar, 1984).