Additives

Additives are used to make tobacco products more acceptable to consumers. They might prolong shelf life (humectants), make the smoke milder and easier to inhale (sugars and humectants), add flavor and aroma, improve the delivery of nicotine (ammonia compounds), and numb the throat (menthol and eugenol). For some time public health workers have pressed for disclosure of additives because some of them, or their combustion products, might be toxic (6). On at least one occasion the tobacco industry has considered additives to reduce the carcinogenicity of cigarette smoke. One such proposed additive, called Chemosol, is discussed in the documents.

By 1981 there was a tension between an increased need for additives in product design, because of pressures to lower tar and nicotine deliveries, and the increased likelihood of government regulation. BAT scientists at the Pichlarn, Austria, research conference in August 1981 concluded:

While additives will assume increasing importance in product design, the freedom to use them could become increasingly restricted. {1178.01, p. 14}

The documents disclose several instances in which B&W was reluctant to abandon the use of additives that independent tests on animals had found toxic. Of course, the company was under no regulatory obligation to stop using these additives; however, it appears to have voluntarily discontinued use of some additives by 1994, probably because of the threat of public disclosure and government regulation.

Chemosol: An Additive To Reduce Cancer

In the late 1960s the American cigarette companies briefly joined forces to sponsor the study of a cigarette additive, Chemosol, which promised to greatly reduce the risk of smoking-induced cancer. The willingness of these companies to come together for such an extraordinary activity speaks eloquently to the seriousness with which the whole industry

privately regarded the cancer risks of cigarettes. In addition, the Chemosol story illustrates the extent to which lawyers controlled critical pathways of scientific endeavor about sensitive subjects within the industry.

Chemosol, described in the documents as a "fuel additive," had been developed under the name "flowebin" by a physician named Max Bindig from Munich, Germany {1807.06}. When used as an additive in cigarettes, it was supposed to promote more complete combustion, which would reduce the levels of the carcinogen benzo(a)pyrene (or benzpyrene) in mainstream smoke. When the rights to flowebin were acquired by a US company, the Chemical Research and Development Corporation, the additive was renamed Chemosol and the company was renamed the American Chemosol Corporation.

During 1966 Dr. Perry Hudson, a biologist affiliated with Columbia University, who ran a research laboratory called the High Tor Foundation, conducted a mouse skin—painting experiment in which cigarettes treated with Chemosol were compared with cigarettes that had not been treated {1807.02}. In this experiment mice exposed to standard tobacco smoke condensate developed a high percentage of malignant tumors, whereas animals exposed to Chemosol-treated tobacco smoke condensate did not develop tumors. This result suggested that Chemosol reduced the carcinogenicity of tobacco smoke condensate.

In the spring of 1967, Joachim Schumacher, vice president of McDonnell & Company, agents for the owners of Chemosol, wrote Ed Finch, president of B&W, about the additive {1807.06}. The letter was a follow-up to a meeting between Schumacher and B&W general counsel, Addison Yeaman. While Schumacher was offering B&W an opportunity to obtain an exclusive license for Chemosol, B&W seems to have been committed from the beginning to bringing this potential additive to the attention of the industry as a whole {1807.07}. The file copy of a letter to the executive director of the Council for Tobacco Research, dated December 12, 1967, states:

Some time ago Ed Finch was approached by representatives of Chemical Research and Development Corporation with an offer to negotiate with Brown & Williamson for an exclusive license to an undefined additive to tobacco called CHEMOSOL . It was claimed that CHEMOSOL I ) very significantly reduced the benzo-a-pyrene content of the mainstream smoke, and 2) that smoke condensate from tobacco treated with CHEMOSOL was virtually free of biologic activity. Finch replied that under no circumstance would Brown & Williamson consider such a proposal, but was willing to act as the conduit of information from the promoters of CHEMOSOL to the United States cigarette industry. {1807.07}

Dr. Griffith reviewed the data on Chemosol that Dr. Hudson had prepared and shared his thoughts in a July 19, 1967, memorandum to Addison Yeaman {1807.01}. Although he was impressed with the quality of the technical report and the soundness of the mouse skin—painting protocol that Dr. Hudson had used, he expressed considerable skepticism about Chemosol. He strongly recommended an attempt at replication of Dr. Hudson's study in another laboratory.

Plans for such a replication apparently were already under way when Dr. Griffith wrote this memo. By August 1, 1967, the Hazleton Laboratory in Falls Church, Virginia (a private laboratory), had submitted a draft protocol for testing Chemosol {1807.08}. Finch's initial decision to involve the entire industry in any development of Chemosol was maintained: nine companies (including B&W, Lorillard, R. J. Reynolds, Philip Morris, and American Tobacco) were sponsors of the 1967 protocol that Hazleton had developed (7). However, the experiment had still not been done by the time the matter was brought to the attention of a congressional committee in April 1969.

During the House hearings on cigarette package labeling in 1969, Dr. Hudson testified about his work on Chemosol. He said that it was capable of reducing the benzo(a)pyrene level in cigarette smoke by 34 percent. He described a mouse skin—painting experiment in which only 5 percent of the mice receiving the condensate from Chemosol-treated cigarettes, compared with 25 percent of those whose skin had been painted with the conventional condensate, developed cancer (7). Dr. Hudson also indicated that scientists in academic medicine and from the National Cancer Institute had reviewed the work done on Chemosol and had found it promising. His striking testimony was covered in the New York Times the following day (8), and by the end of August the tobacco industry was publicizing its agreement with Hazleton to proceed with the experiment (9). It appears that the glare of publicity got this project moving again.

On September 10, 1969, representatives of Hazleton met with the research directors of the major tobacco companies to discuss the protocol. Five days later, another meeting was held with representatives of Hazleton, three tobacco company representatives, and two lawyers from Covington and Burling (counsel for the Tobacco Institute) to discuss the fine points of the proposed protocol. The meeting is documented by a detailed set of notes written the following day by Allan J. Topol, one of the two lawyers present {1807.08}. Topol wrote these notes in the form of a memo to the other lawyer who was there, H. Thomas Austern. Austern

was a prominent expert in food and drug law at the time. He had represented the Tobacco Institute before the Federal Trade Commission a few years earlier (10). The lawyers (purportedly on behalf of the research directors) sought to add defensive language to the protocol. According to Topol's memo,

AJT [Topol] stated that the research directors had decided that the objective [statement in the protocol] should include a sentence stating that the relationship of skin painting to smoking and health had not been established. {1807.08, p. 1}

This embellishment was proposed despite the fact that mouse skin painting was being used to guide product development of less toxic cigarettes within the industry (see chapter 4). The Hazleton spokesman, a Mr. Gargus, disagreed with this change on the grounds that such a disclaimer belonged in the final report, not the protocol.

Although Austern was present as counsel for the consortium of tobacco companies sponsoring the research, he made a number of substantive scientific suggestions about the conduct and organization of the project. He asked whether a sufficient number of cigarettes would be available for the entire project, a question that led to agreement to raise the number to be made from 500,000 to 1 million. He offered the services of the Tobacco Institute Testing Laboratory to conduct preliminary quality control testing of the experimental cigarettes, and he suggested that organoleptic tests (human panel tests of subjective smoking qualities) be added to the protocol. He also suggested that the protocol specify the type of cigarette paper to be used in the manufacture of the experimental cigarettes.

Hazleton had planned to assay the smoke of experimental cigarettes for benzo(a)pyrene levels, but Austern objected

in view of the claims made by Chemosol for benzo(a)pyrene reduction, and in view of the decision of the research directors at the meeting of September 10, 1969, that no decision would now be made as to the test for benzo(a)pyrene content, and that such a decision would be made at the termination of the biological tests. {1807.08, p. 3}

The tobacco company sponsors apparently did not want to consider collecting data that might validate Chemosol's claim for reducing benzo(a)pyrene levels unless the additive was first proven to protect mice from developing cancers. The report indicated that Hazleton's proposed assay of the condensate for benzo(a)pyrene would have been for routine quality control. It is interesting, then, to speculate as to why the attor-

ney chose to object to the use of such a routine procedure. Our hypothesis is that if the experiment showed no protection, but benzo(a)pyrene levels were, in fact, lower, this result would suggest that other carcinogens in the smoke were responsible for cigarettes' carcinogenicity.

Austern also objected to the use of benzo(a)pyrene as the carcinogen to be employed for a positive control in the experiment. It was routine practice to include a positive control in these studies, and benzo(a)pyrene was the one that Hazleton had used in the past. (In fact, BAT later used benzo(a)pyrene—which it described as a "pure carcinogen"—as a positive control in precisely this way in its Janus mouse skin—painting experiments {1138.02, p. 4}; see chapter 4.)

Mr. Gargus explained that benzo(a)pyrene had been employed because it was the best known carcinogenic agent and the one generally used in experiments of this type ... HTA [Austern] suggested that Hazleton employ some control other than benzo(a)pyrene in view of the repeated assertions made by Chemosol for benzo(a)pyrene reduction, and in view of the psychological implications of "benzo(a)pyrene" in the cigarette controversy . Mr. Gargus rejected these suggestions, and Mr. Jessup [of Hazleton] stated that unless Mr. Gargus was satisfied on the scientific grounds, Hazleton would not enter into the contract [emphasis added]. {1807.08, p. 4}

Hazleton retained its preferred positive control, but Austern's attempt to get it changed is an example of how an attorney placed public relations concerns—not the lawyer's traditional concern with liability—ahead of scientific judgments.

Topol's account also describes a substantive discussion about the publication of the results of the experiment. Austern sought to establish control of the results by the tobacco companies (see chapter 8):

With regard to publication of the results, Mr. Gargus suggested that Hazleton be authorized to publish the results of its study in a scientific journal even prior to submission of a report by Hazleton to the sponsors, and prior to Hazleton's submission of its final results to the sponsors. In this way, Mr. Gargus stated, Hazleton could secure its scientific reputation by making sure that Hazleton published its test results. Under this approach no-one would see the Hazleton results until they were published. HTA [Austern] disagreed with this suggestion, stating that the sponsors were paying for the work and should certainly see the results before they were published. In addition, publication is generally a matter of the sponsors' discretion. {1807.08, p. 5}

Austern prevailed. After discussion, Gargus agreed to submit a copy of the report to the nine sponsoring tobacco companies for review prior to publication. However, he insisted that there be an explicit clause in the

contract relating to publication rights. This matter was to be decided later, after Austern had had an opportunity to discuss it fully with his clients.

The following suggestion was also made at this meeting:

HTA [Austern] further suggested that AJT [Topol] should be present at any future technical discussions between Hazleton and industry representatives. {1807.08, p. 6}

In other words, a Convington and Burling lawyer was to participate in all scientific discussions about this experiment between Hazleton and the sponsors.

An interesting facet of the Chemosol story is Ed Finch's insistence that the additive be tested and developed as an industry-wide joint project. Finch may have felt this way for two reasons. In the United Kingdom BAT was part of an industry-wide consortium, the Tobacco Manufacturers' Standing Committee (TMSC), which was committed to the joint development of safe cigarette technology through its Harrogate research lab. Moreover, B&W may still have harbored memories of the reaction felt within the company when Lorillard had gone out on its own in 1962 and developed a filter that selectively removed phenols (see chapter 4). Eight other cigarette manufacturers agreed with B&W's position that Chemosol should be examined jointly, despite the fact that the joint development of a commercially important additive might have invited examination of this activity from an antitrust perspective (see chapter 7).

The documents provide no information on whether the Hazleton protocol was ever implemented or, if it was, what the results may have been. We are not aware that there was any further press coverage, either. There is no indication that Chemosol was ever used commercially.

What is certain is that nine cigarette manufacturers were willing to sponsor research on an additive that promised to reduce the risk of cancer from cigarette smoking; and this venture was coordinated by a lawyer who had worked for the Tobacco Institute. The record of the 1969 negotiation session about the protocol indicates that lawyers played a remarkably large role in discussions about the procedures for carrying out the scientific work. Moreover, there were plans for continued involvement by lawyers in the Chemosol studies at every significant step.

Freon

Tobacco used in cigarettes may be "puffed" or expanded by a variety of processes. Expansion decreases the mass per unit weight, which is one way to reduce "tar" delivery and total tobacco content of a cigarette. In

the 1970s R. J. Reynolds developed a process called G-13, which used Freon-11 for expanding tobacco (11, 12). The process was used by RJR, Liggett & Myers, American Tobacco, and Lorillard. Only Philip Morris and B&W were not using this process as of mid-1977 {1309.01}. An undated and unsigned report, apparently reviewing toxicology studies conducted by RJR for B&W, indicates that in the 1970s RJR tested Freon-11 for its carcinogenicity in mouse skin–painting experiments and also attempted to determine whether Freon had any acute toxicity for the respiratory system {1309.04}. The report indicated that scientists at RJR found no evidence that Freon-11 was toxic under the conditions tested, but the report expressed reservations about these conclusions, since the data had not been evaluated by an independent, outside consultant.

A June 2, 1977, memorandum from Ernest Pepples, the general counsel at B&W, to Dr. I. W. Hughes, head of R&D, describes Pepples's objections to B&W's use of tobacco treated with the G-13 process {1309.01}. Environmental groups had filed petitions with the Consumer Product Safety Commission, and the Environmental Protection Agency was raising concerns that Freon could decompose into the poisonous gas phosgene (which had been used in World War I) when it was burned. The petitioners asked that Freon be banned from air conditioning systems because of this potential hazard. Pepples believed that the finding of phosgene in cigarette smoke might present an opportunity for a products liability claim {1309.01, p. 3}. Moreover, some scientists suspected that Freon could damage the ozone layer of the atmosphere. Pepples worried that the same thing was happening when the Freon residue on the tobacco was heated by the burning coal of a cigarette {1309.06}, although it was not then publicly known that Freon was widely used within the cigarette industry {1309.01, p. 2}:

As we have previously discussed there are three problem areas in the use of the G-13 process:

Pepples also made a strong argument for environmental protection:

There is another factor about which extensive comments are not needed. I refer to the morality of a step which would add 600 lbs. of F-11 [Freon-11] per day to the environment in the face of the already substantial scientific evidence that the chemical is actually harmful to the environment. {1309.01, p. 2}

Pepples recommended that other processes be developed.

A memorandum dated December 5, 1977, from Stanley L. Temoko of Covington and Burling to the Committee of Counsel, a Tobacco Institute committee consisting of the chief counsels of the member companies, shared the news that the Consumer Product Safety Commission had denied the citizen petition "requesting a ban of 'Refrigerant 11' under the Consumer Product Safety Act" {1309.05}. With the threat of regulatory action against Freon diminished, Brown and Williamson began to utilize the G-13 process {1309.02}.

A May 1, 1978, memorandum from J. W. Webb, a research scientist at B&W, discusses the analysis of an expected shipment of expanded tobacco {1309.03}. Tests to be run included moisture content, glycerin content, residual Freon level, yield, percent fines, and fill value. These measures indicate that B&W was evaluating expanded tobacco made elsewhere for possible use. In June 1978 B&W signed a contract with Liggett Group to buy Liggett's excess capacity of G-13 processed tobacco {1309.02}.

Cigarettes With "Nothing Artificial Added"

Cigarettes are more than shredded tobacco wrapped in paper. Additives are required for the tobacco, the papers, and the filter for a variety of purposes. The core need for some additives in modern cigarettes is revealed in a July 1977 memorandum summarizing a meeting at which B&W executives critically examined the claim "nothing artificial added" that RJR was then making about its new brand, Real {1311.01}. Real brand cigarettes had been launched by RJR in 1975, and a menthol version was marketed in 1977. Real contained glycerin (which could have been derived from natural sources) as a humectant; it also contained sugars (and possibly flavorings) for casing the tobacco used in the blend. The menthol version of Real might have contained natural menthol, but that was not yet clear to the B&W group.

B&W could match the RJR claim with a new brand if need be, and it could even use a claim of "all natural flavors." However, the use of "such artificial elements as cellulose acetate, plasticizers, freon, propylene glycol, humectants and some flavors" would preclude the assertion of a "pure organic" claim {1311.01, p. 1}. Propylene glycol, a manufactured chemical commonly used as a humectant by both B&W and RJR, could not be used in a cigarette claiming "nothing artificial added." Tobacco extracts were felt to be "nature identical" at best. Their use in a cigarette would preclude a claim of that brand's being "natural" {1311.01, p. 2}.

The observation that tobacco extracts are not "natural" runs counter to the impression sometimes given by tobacco companies that tobacco extracts are, simply, a form of tobacco. In 1988, when RJR explained its novel nicotine delivery device, Premier, to the scientific community, it described the tobacco extract used in the product as "spray dried tobacco" (13). This phrase was used to suggest that the extract was tobacco, just as certainly as tobacco leaf was. The contrary view expressed privately a decade earlier by B&W scientists in Louisville contradicts this contention.

Eugenol

Eugenol, the active ingredient in cloves, is a local anesthetic agent. It is important for clove cigarettes (cigarettes made from a mixture of cloves and tobacco), which constitute an enormous share of the cigarette market in Indonesia, where BAT has long produced conventional, Western-style cigarettes in competition with clove brands. Traditional clove cigarettes have high tar and high nicotine yields and are made from dark tobaccos instead of the light or blond tobaccos used in Western-style brands. The dark tobaccos produce a harsher smoke, and it is thought that the cloves make it easier for a consumer to inhale the smoke because of the anesthetizing action of the eugenol (14).

B&W was using eugenol as an additive in its cigarettes in 1979 {1822.04}. Eugenol was approved as an additive in cigarettes by BAT's Additives Guidance Panel because of its successful performance in inhalation tests, Ames tests, and other tests from 1982 {1006.01}. The Additives Guidance Panel was an internal BAT group that met in London to review recommendations by its Research and Development Establishment on tobacco additives from various suppliers. Although the government in Jakarta forbade foreign companies from making clove cigarettes, Indonesian clove cigarette manufacturers had begun establishing markets for these products in Australia, Malaysia, and the United States by the early 1980s. This situation afforded BAT a potential market opportunity outside Indonesia for selling clove cigarettes. A summary of the toxicology of eugenol is contained in a 1982 report {1133.01}. This document may be the sort of report on which the Additives Guidance Panel based its decisions.

BAT's interest in eugenol was broader than the Indonesian market, however, and its conception of the pharmacological role of the compound was larger than the local anesthetic action. The BAT research conference held in Brazil in 1983 included a session devoted to eugenol

{1180.28, pp. BW-W2-03741–BW-W2-03742; 1180.07, p. 12}. The discussion centered on developing techniques for measuring eugenol so that its pharmacokinetics and pharmacology could be worked out. Of particular interest to the BAT scientists was its possible action as a central nervous system (CNS) depressant, one that was possibly synergistic with barbiturates and alcohol {1180.28}. The conference summary on eugenol states:

Recognizing the potential for eugenol-flavoured cigarettes in countries outside Indonesia (eg Australia, Malaysia, USA), support was given for the proposed limited GR&DC [Group Research and Development Centre] programme to assess the pharmacology, biochemistry and toxicology of eugenol and its smoke products. The study will include the possible interaction of nicotine as a stimulant with eugenol as a depressant . Significant progress can be expected on this project within one year [emphasis added]. {1180.07, p. 12}

Cocoa

Cocoa has long been an ingredient in American blend cigarettes. In the 1970s the National Cancer Institute's (NCI) cigarette testing program found that cocoa added to a tobacco blend increases the carcinogenicity of cigarette smoke condensate {1319.03}. While the results were not statistically significant, they prompted the British government to ban the use of cocoa in tobacco products. By 1983, however—after additional toxicological data were submitted to the British government's Independent Scientific Committee—the cocoa ban was lifted in the UK for additions of cocoa up to 5 percent of the tobacco weight {1319.03}

An article about cocoa came to the attention of senior executives at B&W in June 1984. The article itself is not in the documents, and it is not clear how it came to B&W's attention. General Counsel Ernest Pepples shared the article with B&W executive vice president Thomas Sandefur, who in turn passed it on to E. E. Kohnhorst, vice president of R&D {1319.02}. Kohnhorst asked Gil Esterle, B&W manager of smoking and health affairs, and Jim Rosene, a B&W research scientist, to analyze it {1319.01}. Esterle and Rosene gave their report to Kohnhorst by early July {1319.03}.

Esterle and Rosene's report confirms that cocoa is used extensively by the industry as an additive and summarizes the NCI findings that cocoa increases the carcinogenicity of tobacco smoke condensate. In B&W products, the report indicates, cocoa levels in cigarettes are about 0.5 percent and up to 5 percent in pipe tobaccos; Philip Morris and R. J. Reynolds

use about the same amounts of cocoa, but Lorillard has replaced cocoa with a proprietary substitute. Lorillard's substitution of something else for cocoa had been verified by the absence of theobromine (an alkaloid in cocoa) from Lorillard brand cigarettes. The document continues:

In the recent past, we attempted with only very limited success to delete cocoa from our cigarettes. Cocoa is incorporated in most of our new products. To eliminate cocoa from all our brands would be a multi-year, costly program, with all indications of minimal success as measured against competitors [sic ] products that use cocoa. Consequently, we recommend not actively pursuing a cocoa deletion/replacement program at this time. Rather the industry should defend the use of cocoa [emphasis added]. {1319.03, pp. 2–3}

A handwritten summary of this report indicates that Kohnhorst shared the main points with Sandefur {1319.01}. So, even though Lorillard had successfully stopped using cocoa in its cigarettes, B&W decided not to do so itself because of economic concerns. The data indicating that cocoa increases the carcinogenicity of cigarette smoke did not seem overwhelming to B&W at the time, and there was no governmental pressure for it to make the change.

Coumarin (Deer Tongue)

Coumarin, which imparts a vanilla-like flavor and is a flavor fixative, was used as a food and tobacco additive for many decades. It was especially valued by tobacco product designers as a flavor booster in low-tar brands. Most commercial use was abandoned when it was discovered that coumarin caused liver damage in both rats and dogs and was suspected of being carcinogenic.

Some background on the B&W attitude toward coumarin is reflected in the January 1978 minutes of an internal BAT Additives Guidance Panel meeting on November 15, 1977 {1310.02}. The minutes indicate that numerous additives were approved for use, and that coumarin received special consideration:

The Hunter Committee [an independent committee in the UK which advised the government on smoking and health] have now considered the Final Report from BIBRA [British Industry Biological Research Association, a report apparently sponsored by BAT (see below)] but were unable to approve levels as high as those recommended by the panel at their last meeting. The reason for this was that the Hunter Committee took into account not only the intake of coumarin from tobacco smoking but also that likely to be ingested from foods in normal diets having a content of natural coumarin. Accordingly,

for the United Kingdom: THE PANEL RECOMMENDS THAT WHERE COUMARIN IS USED AS A TOBACCO ADDITIVE THE RATES OF APPLICATION SHOULD NOT EXCEED

While the limits by the Hunter Committee must be observed in the U.K. the Panel saw no reason to modify their general recommendation of revised limits (740ppm for cigarette tobacco and 1500ppm for pipe tobacco) for other countries, where no legal restriction is imposed. {1310.02, pp. 1–2}

In other words, BAT would observe the UK's prescribed limit on coumarin for products sold in the United Kingdom but not for products sold elsewhere.

On June 6, 1984, E. E. Kohnhorst, sent Thomas Sandefur a memo on coumarin in response to an inquiry from Sandefur {1320.02}. In this memo Kohnhorst indicates that R. J. Reynolds's Now and Vantage cigarettes had coumarin in them in 1982 and 1983, respectively, but the 1984 products no longer did, and efforts were under way to remove coumarin from the "RALEIGH/FIESTA flavor package."

Coumarin is specifically rejected by FDA for use on foods. This rejection was based on very questionable animal studies; but, the FDA has not lifted the ban.

BAT sponsored a study by B ritish I ndustry B iological R esearch A ssociation [BIBRA] on coumarin. The BIBRA study, which is published, showed that coumarin is of no health risk in long-term feeding to baboons.

Presented with the BIBRA results, the UK Independent Committee on Smoking and Health (Hunter, Frogget) lifted this past refusal to allow use of coumarin on tobacco products. This was published in the London Gazette. So, coumarin can be used in the UK and other areas under English influence. BAT (Germany) presented the BIBRA finding to their government decision-makers and were subsequently granted limited approval to use coumarin. This was with low delivery cigarettes (need flavor to induce good smoking quality). There was the requirement to repeat the early studies on coumarin to show they were in error. This accounts for cautious, limited use.

The BAT Additives Guidance Panel (AGP), since the BIBRA findings, has opened [opined?] that coumarin can be used.

BARCLAY originally had coumarin (12 ppm), but it was removed from the flavor formula in November, 1982 . Export cigarettes produced by B&W have not contained coumarin since June, 1983 . We do not sell licensees/associates coumarin or coumarin-containing botanicals [emphasis in original]. {1320.02, pp. 1–2}

Two weeks later, C. J. Rosene wrote General Counsel Ernest Pepples (with a copy to Kohnhorst) about coumarin substitutes {1320.01}:

There are no compounds or formulated flavors available to us that can accurately replace coumarin as a flavor on tobacco. Coumarin itself and natural substances containing coumarin are prohibited from direct addition to food as cited under 21 CFR [Code of Federal Regulations ] 189.130.

Two homologs, 6-methylcoumarin and dihydrocoumarin, exhibit coumarin-like qualities on tobacco but neither can be used as a direct replacement. Both compounds are FEMA-GRAS but are not listed as approved direct food additives by FDA.

We currently use an IFF [International Flavorings and Fragrances, Inc.]-formulated coumarin substitute on BARCLAY . It does not match coumarin very well either.

Attached is a list of domestic brands containing vanillin, mace oil, and glycerin, added either as pure compounds or as part of proprietary flavors. {1320.01}

While FEMA regarded the listed coumarin substitutes as "Generally Recognized as Safe," the FDA did not.

The following month, on July 19, Dr. Sam R. Evelyn from BAT's research laboratory in Southampton sent Kohnhorst a packet of material from his files on coumarin. The material includes reports and site visits to labs conducting animal studies on coumarin as well as a letter from a pathologist at the FDA who had re-reviewed slides from a supposedly positive animal study of carcinogenesis; he had not confirmed the initial report of cancer but had confirmed that coumarin caused liver damage. While the background material Evelyn provided his colleague at B&W is generally negative or noncommittal, his letter indicates that current work with coumarin-feeding experiments was showing some instances of metastatic cancer in the animals. He speculated that such results would likely lead some countries to ban coumarin from tobacco products {1323.02}. Even though the United States had not eliminated coumarin from tobacco as of 1994, it did not appear on a list of cigarette additives released by the six major cigarette manufacturers (2).

Diethylene Glycol (DEG)

Humectants, or moisturizing agents, are used in cigarette tobacco blends to assist with aerosol formation and thus make cigarette smoke "milder." The more that nicotine can be dissolved in the tar droplets, the less irritating the smoke is to the consumer's throat and the easier it is to inhale. Diethylene glycol, more familiar to most readers as an automotive antifreeze, was introduced as a humectant in cigarettes in the 1930s by Philip Morris.

Humectants were reviewed by BAT's internal Additives Guidance Panel in April 1965 {1310.01}. The minutes of the April 9 meeting make the following observations about humectants:

Di-ethylene glycol After some discussion, the Panel members present agreed that because a) there is a known cumulative effect of small doses of this material in the kidneys, and b) there are some other doubtful aspects, it would advise against its use.

Glycerol as a tobacco additive The Panel took the view that, although there was no known reason at present to suppose that this material should not be used on tobacco products, it should, in view of the observations (not supported by references) made on page 62 of the U.S. Surgeon General's Report "Smoking and Health" [the 1964 report] and because of its wide use in the Group, be investigated by R.&D.E. [Research and Development Establishment] both biologically and chemically.

Propylene glycol as a tobacco additive The Panel agreed that this material should also be investigated by R.&D.E. both biologically and chemically. {1310.01, pp. 1–2}

The 1964 Surgeon General's report stated (15, p. 62):

Cigarette manufacture in the United States includes use of additives such as sugars, humectants, synthetic flavors, licorice, menthol, vanillin, and rum. Glycerol and methylglycerol are looked on with disfavor as humectants because on pyrolysis [burning] they yield the irritants acrolein and methylglyoxal. Additives have not been used in the manufacture of domestic British cigarettes since the Customs and Excise Act of 1952, Clause 176, and probably longer, inasmuch as Section 5 of the Tobacco Act of 1842 imposed a widespread prohibition on the use of additives in tobacco manufacture.

Despite the recommendation from the Additives Guidance Panel, B&W was still using DEG twenty years later. In June 1984 R. H. Sachs, deputy general counsel for B&W, wrote to Ernest Pepples, J. G. Esterle, and C. J. Rosene about recent discussions between B&W and Union Carbide, the company that was selling diethylene glycol to B&W {1322.02}. In the course of the discussions, a Union Carbide sales representative said that the company had not known that B&W was using DEG in smoking products and that it would not sell DEG to B&W for this purpose. According to Sachs,

Scientific references were cited to us as the bases for their action. Our scientists read the references, but could not see what Union Carbide was so upset about.

We received a letter from the Union Carbide sales representative stating that DEG could not continue to be used on smoking tobacco because of health concerns, that Union Carbide would be glad to supply a substitute, and that Union Carbide only recently found out about our use of DEG on smoking tobacco.

I contacted the Union Carbide General Counsel to discuss the matter. After an investigation, a lawyer in his office called me to say:

They are willing to cooperate in any way they can to rectify this situation.

While Union Carbide is satisfied that our continued use of DEG in smoking tobacco does not pose a health hazard, we should not let the matter rest there. If the "ingestion" of DEG applied to chewing tobacco is a problem (as Union Carbide seems to indicate it is), then how is it that they are so sure that its inclusion in smoking tobacco poses no problem?

We will look into this matter further [emphasis in original]. {1322.02, pp. 1–2}

The memo includes handwritten notations indicating that Earl (Kohnhorst) asked Gil Esterle whether the effort to eliminate DEG was continuing. Esterle replied that it was. He noted that DEG was not "Generally Recognized as Safe" by either the FDA or FEMA, and that it was not used in foods.

[I]f we need to someday release to government we will need to vigorously defend and public reaction would be negative (I don't trust confidential disclosures to government). We should replace but with a good match. {1322.02, p. 1}

The list of additives released by the tobacco industry in April 1994 does not include DEG (2).

Public relations seems to have become the guiding force for managing potentially toxic additives such as DEG, as it finally became for dictating how to handle matters related to the toxicity of tobacco (see chapter 4). The operative approach was that ingredients for cigarettes must be proven dangerous, and an acceptable substitute be available, before changes would be made. The usual expectation for consumer products, however, is that safety should be established before a product is used.