Criteria for development of a database for safety evaluation of fragrance ingredients.

Over 2000 different ingredients are used in the manufacture of fragrances. The majority of these ingredients have been used for many decades. Despite this long history of use, all of these ingredients need continued monitoring to ensure that each ingredient meets acceptable safety standards. As with other large databases of existing chemicals, fulfilling this need requires an organized approach to identify the most important potential hazards. One such approach, specifically considering the dermal route of exposure as the most relevant one for fragrance ingredients, has been developed. This approach provides a rational selection of materials for review and gives guidance for determining the test data that would normally be considered necessary for the elevation of safety under intended conditions of use. As a first step, the process takes into account the following criteria: quantity of use, consumer exposure, and chemical structure. These are then used for the orderly selection of materials for review with higher quantity, higher exposure, and the presence of defined structural alerts all contributing to a higher priority for review. These structural alerts along with certain exposure and volume limits are then used to develop guidelines for determining the quality and quantity of data considered necessary to support an adequate safety evaluation of the chosen materials, taking into account existing data on the substance itself as well as on closely related analogs. This approach can be considered an alternative to testing; therefore, it is designed to be conservative but not so much so as to require excessive effort when not justified.

A parallelogram approach for safety evaluation of ingested acetaldehyde.

Route-specific carcinogenicity data are often lacking for compounds of regulatory importance. Acetaldehyde (AA), for example, a natural constituent in foods, is a rodent carcinogen via the inhalation route, but oral carcinogenicity data are not available. In the absence of such data, a parallelogram approach can be used to estimate the oral carcinogenic potency of this chemical. The relative potency of AA to the structurally related compound formaldehyde (FA) in the nose and the relative potency of FA in the nose and stomach serve as the basis for estimating the potency of AA in the stomach. On a dosimetric basis, inhaled AA is 14- to 35-fold less potent than FA in producing nasal DNA-protein crosslinks (DPX), subchronic tissue injury, and tumors. Ingested AA is also considerably ( approximately 5-fold) less potent than FA in producing gastric injury and DPX. Compared to the nose, the stomach is 10- to 60-fold less sensitive to both AA- and FA-induced DPX and subchronic tissue injury. The parallelogram approach will not supplant long-term oral carcinogenicity studies with AA; however, the consistent pattern of decreased sensitivity of acetaldehyde compared to formaldehyde, lower sensitivity of the stomach to the nose, and the lack of gastric tumorigenicity of orally ingested formaldehyde strongly suggests that ingested acetaldehyde is not likely to be carcinogenic. Successful estimation of the carcinogenic potency of ingested glutaraldehyde, for which chronic oral and inhalation data are available, provides further support that the parallelogram approach can provide a reasonable estimate of the carcinogenic potency of closely related aldehydes, such as AA.

Toxicological evaluation of commercial alkyldimethylamine oxides: two-year chronic feeding and dermal studies.

Chronic feeding and dermal studies were conducted with commercial alkyldimethylamine oxides (ADAO) in rats and mice. Rats were dosed with 0.01, 0.1, or 0.2% w/v ADAO (on a 100% active basis) in their diet for 104 weeks. Mice were exposed to ADAO via a dermal application of 0.1 ml of an aqueous solution of ADAO at concentrations of 0.05, 0.13, or 0.26% w/v ADAO (on a 100% active basis) once daily three times per week for 104 weeks. No compound-related carcinogenic effects were observed in either chronic test.

Differential effects of Walker 256 carcinosarcoma cells growing subcutaneously, intramuscularly, or intraperitoneally on hepatic microsomal mixed-function oxygenase activity.

Walker 256 rat carcinosarcoma cells growing as solid subcutaneous or intramuscular tumors depressed hepatic microsomal mixed-function oxygenase activity to less than 20% of control activity, but the same tumor cells growing as free ascites cells in the peritoneal cavity did not. Necrosis of the core area of solid tumors was observed. Tumor cells may release a substance that depresses hepatic microsomal mixed-function oxygenase activity only upon their death.