Tobacco Documents Online

A PROPOSAL BEFORE THE SAFETY EVALUATION AND COMMITTEE (SEAC) )/ A combination of flavoring ingredients representing a prototype new product recipe (Harley-Davidson Lights) was submitted by the Product Development Department for review by SEAC. The Committee was asked to render an opinion regarding a proposal by Product Development to proceed with commercialization of the new product before the completion of advanced toxicology testing. Background information having potential relevance to this judgement is itemized below and supporting documentation is attached where appropriate. Each member of the committee is requested to review the information and statements below and to determine whether the statements are, in the member's own judgement, reasonable based upon available scientific data. Members are invited to edit, revise, or propose alternatives to the statements below. SEAC will meet in the very near future in an attempt to forge a consensus judgement and advisory opinion to address the request from Product Development. 1. THE PROTOTYPE NEW PRODUCT RECIPE IS SUBSTANTIALLY SIMILAR TO SEVERAL EXISTING RECIPES WHICH HAVE BEEN THOROUGHLY TESTED. The new recipe consists of sugars and humectants, which are exempt from testing, and 8 flavoring ingredients (TABLE 1). Three of these flavoring ingredients (licorice, roasted St. John's bread, menthol) are present in many other marketed recipes at similar or higher levels and have been repeatedly subjected to the advanced toxicology tests comprising Phase 5 of the Ingredients Testing Program (mouse skin painting bioassay and a 13-week subchronic smoke inhalation study) Together, these familiar components comprise 89.99% of the applied ingredients. These major ingredients have a long history of extensive use throughout the industry. Decades of experience has revealed no indications of undesirable interactions among these materials and other flavoring ingredients. The likelihood that these materials might behave in some unique and unfavorable manner in the proposed recipe seems remote. 2. THE PROTOTYPE NEW PRODUCT RECIPE CONTAINS A NUMBER OF COMPONENTS WHICH ARE PRESENT IN MINOR CONCENTRATIONS AND THE LIKELIHOOD THAT THESE COMPONENTS MAY EXHIBIT ANY MEANINGFUL BIOLOGICAL EFFECTS IS EXTREMELY REMOTE Minor components of the recipes include four top flavor materials: linalool, benzaldehyde, anethole, and 2,3,5,6- tetramethylpyrazine. These flavor constituents are applied to

,` .J V , `. 14 the tobacco blend at trivially low levels (TABLE 1), and comprise less than 0.0004%,of the applied ingredient mixture. The combined application rate for all of these ingredients together is less than 0.4 ppm tobacco. For the sake of comparison, the USDA import tolerance for residues of DDT, DDE, and TDE on tobacco is also 0.4 ppm. All four of these minor top flavoring ingredients have been identified as natural leaf components and/or smoke components. All four of these ingredients have completed the entire sequence of testing required for individual ingredients (Phases 1-3) by the Lorillard Testing Program (Attachments 1- 4). 1 Anethole and benzaldehyde are present in other currently 1` ~ marketed product recipes as components of compounded ' flavorings. Benzaldehyde was tested in a previous 13-week cigarette smoke inhalation bioassay_cvs-a1component of a ' compounded flavoring applied at ~30 ppm obacco. b~~7u j~F ~Jq: 1ha ~!qUd / j 3. COCOA, WHICH COMPRISES 1096 OF THE PROPOSED PRODUCT RECIPE, HAS NOT BEEN USED IN ANY OF THE COMPANY'S PRODUCTS IN RECENT YEARS AND HAS NOT BEEN SUBJECTED TO ANY PHASE 5 TESTING. Cocoa is a component of the proposed recipe at a level of 8796 ppm tobacco. While this application rate renders cocoa a major ingredient of the product, this level is reportedly lower than is found in many competitors' products on the market. Cocoa is reportedly in extensive use in the tobacco industry. However, it was removed from the company's products some years ago following a report from the National Cancer Institute's - Tobacco Working Group (NCI) project which described an apparent increased mouse skin tumorigenicity for condensate from cigarettes containing cocoa compared to appropriate reference cigarettes. The removal of cocoa preceded the initiation of the Ingredients Testing Program, so no recent toxicology on cocoa or cocoa-containing recipes has been done by the company. However, a substantial toxicology literature for cocoa and cocoa components is extant. A review of the available cocoa toxicology database reveals no indications of substantial biological activity relevant to tobacco industry uses except for the aforementioned NCI mouse skin painting study. A second mouse skin painting bioassay was published in 1990 which found no increased tumorigenicity by cocoa used as a cigarette ingredient (Attachment 5). This study was the subject of a brief memo by J. D. Heck (Attachment 6) which discussed the more recent study. Unfortunately, the specific type of cocoa used in the skin

painting study was not fully described in terms of fat content. A recent communication with the sponsors of the study (F. Daylor, Philip Morris, Inc., 5/11/93) indicated that the cocoa used in the study was representative of the material commonly employed in tobacco products. However, exact specifications of the cocoa have not been received. SEAC RECOISIENDATION : The major ingredients of the proposed recipe, (humectants, sugars, licorice, roasted St. John's Bread, and menthol), by virtue of their long history of uneventful use in tobacco products and their substantial company toxicological database, may reasonably be anticipated to pose no significant risk in their intended use. The top flavoring constituents of the proposed recipe (linalool, benzaldehyde, anethole, and 2,3,5,6 tetramethylpyrazine~ are naturally present in tobacco, have substantial toxicology data available, and are added at sufficiently low levels to permit a reasonable judgement that they pose no significant risk. Cocoa has a long history of use in the tobacco industry and has a generally favorable toxicologic profile. However, results of two mouse skin painting bioassays have been inconsistent with regard to potential enhancement of cigarette condensate tumorigenesis by cocoa. It is the opinion of the committee that the weight of more recent evidence suggests that the use of cocoa in tobacco products is not likely to contribute to increased toxicity of those products. Commercialization oposed recipe may proceed as long as the Phase 5 tests quired y the Lorillard Testing Program are conducted immediate completed as soon as possible. ~

TABLE 1 HARLEY DAVIDSON LIGHTS Name ppm Testing Status Brown Invert Syrup 20000 Exempt High Fructose 18003 Exempt Propylene Glycol 13999 Exempt Glycerine 10776 Exempt Savannah Cocoa 8796 Not tested Licorice 3602 Skin painting and inhalation up to 21000.ppm St.John's Bread 1203 Skin painting and inhalation up to 13000 ppm Natural Menthol 70.5 Skin painting up to 635 ppm, Inhalation up to 156 ppm Linalool 0.168 Phase 3 complete Benzaldehyde 0.146 Phase 3 complete, Component of PD071 tested in inhalation up to 30 ppm Anethole 0.015 Phase 3 complete 2,3,5,6 tetramethylpyrazine 0.008 Phase 3 complete VzsSzsLe

ATTACFII4ENT 1

LINALOOL TOXICITY PROFILE Linalool (B105) is a naturally occurring compound found in more than 200 oils from herbs, leaves (including tobacco), flowers, and wood. REGULATORY STATUS FEMA GRAS FDA GRAS HUNTER LIST (Not to exceed 0.15% in cigarette, cigar or hand-rolling tobacco and 0.5% in pipe tobacco). JECFA (ADI 0.5 mg/kg) COUNCIL OF EUROPE (0.5 mg/kg for total linalool and linalyl esters) METABOLISM Linalool appears to be readily absorbed after oral administration with excretion of 93% (urine/55%, feces/15%, and expired air/23%) occurring within 72 hours in male Wistar rats (Parke et al, 1974). Linalool has been shown to increase hepatic microsomal enzyme activity in rats (Parke and Rahman, 1969; Parke et al., 1974b; Chadha and Madhava Madyastha, 1984) . A short-term (3 days) study indicated that linalool induces the enzymes involved in its metabolism. A longer study (64 days) demonstrated that effects on liver enzymes develop slowly and may represent a physiological adaptation to linalool exposure. Another study showed that Cytochrome P450 activity increased after 3 days of oral dosing, but decreased to control levels after 6 days of dosing. Cytochrome b5 actitivity exhibited a moderate increase. Roffey et al. (1990) described linalool as a relatively weak peroxisome proliferator. However, no accompanying P450IA1 induction was detected in this assay. Linalool did not induce hepatic enzymes in a screen in Sprague- Dawley rats at 375 mg/kg (unpubl.) ACUTE AND SUECHRONIC TOXICITY Species/Sex Route LD50 Values Mouse/male ip 340 mg/kg (Atassanova-Shopova, 1973) Mouse/UK im 8000 mg/kg (Northover, 1962) Rat/male ip 307 mg/kg po Rat/M,F oral M >464<681 unpubl. F 700 mg/kg C 687 mg/kg po po Rat/UK Rabbit/UK oral dermal 2790 mg/kg 5600 mg/kg (Jenner, 1964) (Opdyke, 1979)

Maximum Tolerated Dose (MTD) Species/Sex Route MTD Mice/M,F ip 125 mg/kg (Stoner, 1963) Rat/M,F oral M 750 mg/kg; F 3750 mg/kg (unpubl. ) Twelve Week Feeding Study Species/Sex Dose level Results Rats/M,F 50 mg/kg/day No adverse effects except slight growth retardation (FEMA, 1975) IRRITATION DATA Linalool has been characterized as moderately to severely irritating to rabbit skin, but non-irritating and non-sensitizing in human subjects (Opdyke, 1979; RTECS, 1988, Fujii et al., 1972 and Greif, 1967). Thelestam et al., (1980) reported a high degree of membrane damage in linalool-treated human lung fibroblasts in vitro. Cell cultures were incubated with a 25 mM solution of linalool in Tris-buffered saline. The degree of membrane damage was was classified as high (>70% nucleotide release), moderate (70 - 45%) and nil (>15%). MUTAGENICITY/GENOTOXICITY ASSAY RESULTS TEST SYSTEM Activation w I w/o Ames Test - - Salmonella typhimurium strains TA92 TA1535, TA100 TA1537 (Ishidate,1984) Ames Test (mod) - - Salmonella typhimurium strain TA100 (Eder, 1980) Ames Salmonella/ - Salmonella typhimurium strains TA100 -microsome and TA98 (Rockwell and Raw, 1979) Ames Test - - Salmonella typhimurium strains TA100

Mouse Lymph. Eq - TA98, TA1535, TA1537 (unpubl.) Mouse Lymphoma cells (unpubl.) UDS negative Rat Primary Hepatocytes (unpubl.) Chr. Aberr. - Chinese hamster fibroblast cells (Ishidate, 1984) Chr. Aberr. - - Chinese hamster ovary cells Sister Chr. Exch. negative (unpubl.) Chinese hamster ovary cells (modified) (Sasaki,1989) CARCINOGENICITY Species/Sex/Route Dosage/Regimen/Duration RESULTS Mouse/M,F/ip 3.0 (MTD) or 0.6 (0.2 MTD)g/kg Negative 3x weekly/8 wks. Evaluated at week 24. Special Studies - Linalool elicited negative results in a study which examined the hepato-carcinogenic potential of various compounds based on liver regeneration rates in partially hepatectomized rats. Linalool-treated animals exhibited comparable liver regeneration rates as compared to control values following subcutaneous administration of 2450 - 2950 mg/kg/day linalool for 7 consecutive days. (Gershbein, 1977). Linalool-treated animals exhibited a reduced level of papillomas when compared to an acetone control group in a modified mouse skin initiation/promotion assay (Gould, 1987). Roe and Field (1965) reported that linalool as a 20% solution in acetone, elicited a weak tumor-promoting response. No data was offered in support of this statement. IMMiJNOMODULATORY Linalool did not suppress the primary antibody response to sheep

red blood cells in B6C3F1 mice at 375 mg/kg. A slight increase in SRBC response was elicited by a dose of 188 mg/kg in mice in another study (unpubl.) ACUTE CARDIOVASCULAR AND RESPIRATORY Linalool did not induce physiological impairment of cardiovascular and pulmonary function following intravenous administration of 0.4, 0.8, and 2.0 mg/kg to female Beagle dogs (unpubl.). PHARMACOLOGICAL STUDIES Linalool was reported to have a vasodilatory effect on blood vessels in the anesthetized dog (Northover and Verghese, 1962). A protective effect on pentylenetetrazol-induced convulsions was attributed to treatment with linalool at 150, 175, and 200 mg/kg in mice and 200 and 300 mg/kg in rats (Atanassova-Shopova et al., 1973) . Linalool has been proposed as an antineoplastic and antiviral agent for veterinary administration (Opdyke, 1979).

ATTACHMENT 2