Quantitative Risk Assessment of Carcinogenicity of Urethane (Ethyl Carbamatc) on the Basis of Long‐term Oral Administration to B6C3F1 Mice

• Published in: Japanese Journal of Cancer Research 1991-04, Vol.82 (4), p.380-385
• Main Authors: Inai, Kouki, Arihiro, Koji, Takeshima, Yukio, Yonehara, Shuji, Tachiyama, Yoshiro, Khatun, Nasima, Nishisaka, Takashi
• Format: Article
• Language: English
• Subjects: Adenoma - chemically induced; Animals; B6C3F1 mouse; Biological and medical sciences; Carcinogenicity; Carcinogens, Environmental; Carcinoma - chemically induced; Chi-Square Distribution; Dose-Response Relationship, Drug; Hemangioma - chemically induced; Hemangiosarcoma - chemically induced; Liver Neoplasms - chemically induced; Logistic Models; Lung Neoplasms - chemically induced; Male; Medical sciences; Mice; Mice, Inbred Strains; Models, Theoretical; Tumors; Urethane; Urethane - administration & dosage; Urethane - toxicity; Virtually safe dose
• ISSN: 0910-5050; 1349-7006; 1876-4673
• DOI: 10.1111/j.1349-7006.1991.tb01859.x

A carcinogenicity study of urethane was performed for quantitative assessment of Its risk in humans. Three hundred 6‐week‐old male B6C3F1 mice were divided into 6 groups, each consisting of 50 mice, and urethane was given ad libitum in drinking water at levels of 0 (control), 0.6, 3, 6, 60 and 600 ppm for 70 weeks. The tumors with a clear dose‐response relationship were lung tumor (alveolar/ bronchiolar adenoma or carcinoma) and liver tumor (hemangioma or angiosarcoma). The incidences of these two types of tumor were applied to estimation of the virtually safe dose (VSD) at the level of 10 6 by using four mathematical models (Logit, Probit, Weihull and Multistage models). The VSD based on the incidences of lung tumor by using the Logit model was estimated to be 1.8 × 10 4 mg/kg body weight/day. On the other hand the VSD based on those of liver tumor by using the Weibull model was 7.2 × 10−5 mg/kg body weight/day. Thus, the VSDs based on the incidences of the two different types of tumor using the most compatible mathematical model in each case, as judged from the P‐values, were similar.