Southampton Research Conference, 1962

As mentioned in chapter 3, BAT held periodic research conferences so that its scientists from around the world could meet to discuss research developments and other issues facing the industry. Such a conference was held at Southampton in early July 1962. The documents include a detailed set of minutes from a major session that dealt with a wide range of smoking and health issues. These included reaction to the recently issued report of the Royal College of Physicians on smoking and health, the company's research program, and research that was being undertaken through an industry-wide consortium in the United Kingdom known as the Tobacco Manufacturers' Standing Committee (TMSC; see chapter 2). As discussed in chapter 2, the British Royal College of Physicians issued its first report on smoking and health in the spring of 1962. That report concluded that cigarettes cause lung cancer and recommended that government take "decisive steps' to control the rising consumption of cigarettes.

The Southampton conference was attended by twenty-nine executives and research scientists from BAT offices and laboratories in five countries (including the United States and Canada, which sent four delegates each) {1102.01, front matter}. The record of the conference shows that, despite its public statements contending that the evidence on smoking as a cause of lung cancer was inconclusive, the British tobacco industry was moving aggressively to examine the problem with sophisticated toxicological testing.

In addition, a special session was devoted to a major problem that had arisen in the United States about six weeks earlier, when a B&W competitor announced its development of a cigarette filter that selectively removed phenols, a group of compounds suspected of enhancing the carcinogenic activity of cigarette smoke.

The documents include a detailed set of minutes from a major session dealing with a wide range of smoking and health issues. In a keynote speech Sir Charles Ellis, an executive in BAT's Research and Development Establishment at its Millbank headquarters in London, presented an overview of the smoking and health issues faced by the company and explained how the TMSC was responding to these issues. The full text of his remarks is included in the minutes. (The parts of his comments related to nicotine were discussed in chapter 3.) Sir Charles reviewed public and scientific perceptions of the cancer issue with a variety of rationalizations, but he also outlined a well-planned, comprehensive research program designed to reduce the toxicity of cigarettes.

Lung Cancer And The British Royal College Of Physicians Report

In his speech Sir Charles tries to downplay the Royal College of Physicians' conclusion that cigarettes cause lung cancer, calling it an emotional reaction based on an incomplete examination of the data. He does, however, acknowledge that smoking causes bronchitis.

After reading the Report of the Royal College of Physicians and the debate in the House of Lords the dominant impression I received was that of people who had reached an emotional conclusion in which they believed passionately and sincerely. This report provided the occasion for statements of faith by people who seemed to find it necessary, however, to silence their own self-criticism by repeating phrases like, "conclusive proof beyond the shadow of doubt" [,] ... "devastating effect of the marshalling of cold scientific facts", and so on. Yet we who have been immersed in the subject for many years know that this report produced no new fact, produced no new arguments, indeed, except for the contribution of an emotional gloss, left the subject untouched. We know only too well that there are no conclusive proofs; that there are few, if any, cold scientific facts.

However, emotional conclusions cannot be disregarded. They may not be right, but they are not necessarily wrong. Emotional judgments are often the basis for national thinking, and since a national attitude to smoking may be building up it is essential for us to consider what are the components in this emotion.

The most important is the dread word "cancer". Most people cease to be able to reason once it is mentioned, and you will all be aware how difficult it is for doctors to overcome the reluctance of people to admit the possibility of having cancer and to present themselves for early examination. Lung cancer carries with it all of these associations, and also shares some of the aura of dread connected with tuberculosis. I can well remember how pneumoconiosis in the coal mines had much of this emotional background and was correspondingly difficult to deal with in a rational manner. Smoker's cough is a real phenomenon and obvious to everyone , and we should recognize that it is a factor in the emotional build-up [emphasis added]. {1102.01, pp. 3–4}

"Smoker's cough" is a symptom of lower airway irritation caused by cigarette smoke. "Bronchitis" is the clinical term that is implied by the phrase "smoker's cough." This irritation occurs as a result of smoke inhalation, which, in turn, is necessary for nicotine absorption from a cigarette (see chapter 3). Where there is a smoker's cough, there is some degree of bronchitis.

Sir Charles states that the evidence directly implicating cigarette smoking as a cause of lung cancer was epidemiological except for a single study by Blacklock (4). Blacklock had injected cigarette smoke con-

densates mixed with the adjuvant eucerin, a material to amplify the effects of smoke condensate, into the lungs of rats, guinea pigs, and rabbits. The animals lived out their natural lives and then were examined. Cancers were observed in six of seventy-two rats and in one rabbit.

In saying that Blacklock's work was the only biological study in the literature that addressed the question of smoking as a cause of lung cancer, Sir Charles is framing the issue very narrowly. A decade earlier, the young US physician Ernst Wynder had shown that cigarette smoke tar condensate caused cancer in a mouse skin model (5, 6). The test involved repeatedly painting smoke condensate onto the skin of mice and observing the animals over a long period of time to see whether cancers developed at the exposed site. Although Sir Charles does not mention the mouse skin–painting model as relevant to the question of whether smoking causes lung cancer, BAT was to use this very model in Project Janus (discussed below) as its gold standard for determining whether specific product modifications could reduce the dangers of smoking.

Joint Research by British Tobacco Industry

Sir Charles then describes the new tobacco toxicology laboratory at Harrogate, which was to be run jointly by tobacco companies in the United Kingdom through the Tobacco Manufacturers' Standing Committee (TMSC). In the course of his discussion, Sir Charles explains the BAT board's policy on the smoking and health issue:

The Board recognises that this problem must be tackled from two sides, the first being medical research on the origin of lung cancer and bio-assay on the biological effects of smoke, and the second being the composition of smoke and the possibilities of modifying it. The Board has decided that if this Company makes any significant scientific discovery clearly relevant to health it will share its knowledge with its co-members of TMSC and not seek to obtain competitive commercial advantage. The Board has therefore decided that they will whole-heartedly support TMSC to carry out and co-ordinate all research on smoking and health. TMSC will do this by itself carrying out biological work at its establishment at Harrogate and by sponsoring biological and medical work at Institutions. TMSC will depend on member companies for physical and chemical work [emphasis added]. {1102.01, p. 5}

While the BAT board made a commitment to share information on how to reduce cigarette toxicity with other tobacco companies rather than "seek[ing] to obtain competitive commercial advantage," there is no mention of making research information available either to the public or to public health authorities.

More important, as of July 1962, BAT was committed to cooperative investigations and product development when they were "health-oriented." A "health-oriented" product, as it would later be defined at the Hilton Head research conference in 1968, was one that reduced the actual hazard of smoking. In contrast, a "health-image" cigarette only appeared to provide a health benefit (see below). The fact that the company had decided to share any promising results with the industry as a whole strongly indicates that the company now realized that its products did, in fact, present a hazard. If the company regarded the evidence that smoking causes cancer as false, misleading, or inconclusive, it is difficult to see why it would have seen the possibility of a competitive advantage and adopted the policy that it did.

Sir Charles couches the rationale for conducting "health-oriented" research in the following guarded terms:

The central fact in this subject is that in sufficient doses tobacco condensate acts as a carcinogen when painted on the backs of mice, or when injected subcutaneously into rats. In sufficient dose it also acts as a co-carcinogen in mouse painting tests. On present evidence the amount of the known carcinogens in smoke are insufficient to make it plausible that these experiments could be extrapolated to support the view that smoking is harmful to human beings, but at least they serve to indicate a group of substances which require intense investigation and which, even if we do not know why, we would be pleased to see less of [emphasis added]. {1102.01, p. 7}

Despite this cautious attitude toward the implications of the mouse skin model, mouse skin painting had by this time become the test used by BAT to examine whether a particular manipulation of cigarettes reduced the risk of cancer. Indeed, discussion later on at this research conference demonstrated the hard-nosed attitude within R&D at BAT that the company would rely primarily on mouse skin tests in estimating the value of any maneuver to remove toxins from cigarette smoke. Mouse skin painting was, then, the definitive assay for cancer at BAT—even though, in public, the tobacco industry minimized the value of data from mouse skin–painting experiments. The public posture was that this model was not at all relevant to a determination of whether cigarettes cause cancer in humans.

The TMSC laboratory, which was to open at Harrogate on September 1, 1962, was to be capable of running mouse skin–painting experiments at twenty-five to fifty times the size of in-house efforts. Two thousand mice were to be used in the initial experiment, instead of the forty to eighty that were conventionally used. The start-up costs were put at

£250,000, and the annual budget was estimated to be £100,000. Research results of this and other studies were to be published as a matter of policy {1102.01, p. 33}. In determining the sample sizes to be utilized in the initial experiment, the planners had done calculations of how small a difference could be detected between the activities (i.e., carcinogenic potential) of different condensates. Sir Charles notes that the experiment at Harrogate "should with almost certainty show up a difference of 30 per cent in the two condensates and will give good information about possible differences of 15 per cent" {1102.01, p. 9}.

One of the criticisms of earlier work was that the smoke condensates used were not "fresh"; that is, they had been stored for some time prior to being painted onto the mice. People inhale "fresh" smoke, not old smoke, and any chemical changes that made the stored smoke more toxic to mouse skin might skew the results. So the first experiment conducted at Harrogate was to be a comparison of "fresh" and "stored" condensate.

Besides looking at carcinogenic and cocarcinogenic activity in smoke, Harrogate was also charged with investigating the mechanisms whereby tobacco smoke causes irritation of the respiratory passages. A cocarcinogen is a material that amplifies the action of a carcinogen but is not itself carcinogenic. This research was relevant to the work at Harrogate, rather than simply a matter for competitive improvement of consumer acceptance, for three reasons. Sir Charles notes that irritation could

The term "chronic bronchitis" is a specific diagnosis of some importance. In 1962 British investigators of what we now call chronic obstructive pulmonary disease (COPD) called the disease "chronic bronchitis," while investigators in the United States used the term "emphysema." The term "COPD" was coined to encompass both major phenomena of the condition, the narrowing of airways and the destruction of lung tissue, so that the literature would be easier to follow. Nearly all COPD seen in the United States is caused by cigarettes (7).

At Harrogate researchers would be able to conduct two kinds of tests: (1) by looking at the proliferation of mucus-secreting goblet cells in the lung (a response to irritation), they could determine the potential of smoke to produce bronchitis; (2) by looking for hyperplasia of the epithelial cells lining the bronchi, they could determine the presence of carcinogenesis

and cocarcinogenesis (the stimulation of these cells to divide is an early action of carcinogens). Harrogate was to have the capability to set up both of these experimental models and then to test specific smoke constituents in these systems.

Sir Charles emphasizes the importance of the studies on irritation with the following comments:

I regard the attempt to probe more deeply into the cause of irritation as equally fundamental [to investigating the cancer problem]. Some smoke is irritating, smoker's cough is a reality, and it cannot be good for health to cause this irritation whether or not this irritation has any effect on the incidence of lung cancer . Some of you may regard irritation as so important that you are willing to make a guess and attribute it with but little evidence to a class of compounds such as aldehydes, or you may be ready to take a short cut and attempt to identify irritation by an organoleptic technique [subjective report of people exposed to potentially irritating materials]. This is a matter on which I hope you will express your views. Personally I think the TMSC policy is right; this subject of irritation is just too important to incur the danger of reaching a wrong conclusion due to a faulty technique. Once we have solved the problem of establishing a quantitative test[,] progress in identifying irritating components will be rapid. Whether their removal will be easy is another question, but to identify them will be a good first step [emphasis added]. {1102.01, p. 22}

Sir Charles also describes other work supported by TMSC outside the Harrogate facility: an examination of the relative carcinogenicity of smoke from tobacco that has been cured in different ways; a large, prospective epidemiological study on susceptibility factors (other than smoking) for cardiac and pulmonary disease among 200,000 men; a study of the carcinogenicity of aliphatic lactones (material found in plants that was thought likely to occur in cigarette smoke); an epidemiological study of lung cancer in South Africa and Australia; and the psychological work of Dr. Hans Eysenck. (Dr. Eysenck later received support through the special projects of the Council for Tobacco Research; see chapter 8.) TMSC also funded work on viruses and immunogens as causes of cancer and on the effects of tumor transplants treated in various ways in experimental animals. Taken as a whole, this work plan has a defensive tone: most of these projects sought explanations for the diseases at issue other than that they are caused by cigarettes.

After describing the overall research activities sponsored by TMSC, Sir Charles returns to the experiment on fresh versus old condensate.

[If] fresh condensate has a smaller biological effect than old condensate we do not of course at one stroke destroy the emotional conclusion [that smok-

ing causes cancer]—in fact it remains unaltered—but we do clear the decks of a great deal of previous experimentation and smoke condensate will then be accepted as such a weak carcinogenic agent that other explanations of the association of smoking and lung cancer assume greater importance. Conversely, and this is always a possibility, the biological effect may increase as the condensate is used fresher and fresher. This possibility need not dismay us, indeed it would mean that there really was a chemical culprit somewhere in smoke, and one, moreover, that underwent a reaction fairly quickly to something else. I feel confident that in this case we could identify this group of substances, and it would be worth almost any effort, by preliminary treatment, additives, or filtration, to get rid of it. We should have brought the problem out into the open where it could be attacked. I feel sure you will give your full support to this experiment [emphasis added]. {1102.01, p. 21}

Sir Charles's willingness to accept the possibility that something intrinsic to cigarette smoking could cause cancer—and that, once identified, this "group of substances" could be removed—reflects a greater level of scientific openness than the tobacco industry has evidenced in recent years.

Ultimately, the experiment showed that "fresh" condensate (twenty-four hours old) actually had more carcinogenic activity than "old" condensate (one week old) (see below) {1105.01}.

The Bat Research Program

BAT's intramural research activities, Sir Charles explains, are devoted mainly to a detailed investigation of the physics and chemistry of tobacco smoke.

Now, as regards work in our own laboratories on chemistry and physics. Our objective must be to prepare ourselves as fully as possible to utilize any result the moment it appears from these biological and medical experiments. In effect, this means that we would like to know the origin of all the compounds that we currently think may be important, whether they are wholly or partially distilled from the tobacco or formed in some of the pyrolytic reactions, and at what temperatures. We might then hope to be in the position of being able to enhance or suppress certain classes of compounds by either pre-treatment or additives once we knew which compounds had to be so influenced.

This of course is precisely what we have been doing and are doing, and in my opinion progress is excellent. {1102.01, p. 24}

These comments indicate that a goal of R&D at BAT in 1962 was to reduce the toxicity of its cigarettes as rapidly as possible.

Sir Charles concludes his conference remarks with the following analysis and exhortation:

We must admit that the threat to our industry is serious and very real, and it is of little help to us that it is based on an emotional guess and not on reasons. I believe we are now starting on a sound programme of investigation that in a few years will make it possible to see the situation and judge the future much more clearly. I hope to-day's discussion will start your active participation in this particular section of Group Research, and that the measures we propose will enable it to continue and grow. {1102.01, p. 25}

Unfortunately, this optimistic view that cigarette toxicity could be largely eliminated through research and engineering proved to be wrong.

Participants at the conference also discussed the feasibility and development of a safer cigarette. It emerged that BAT knew how to make a cigarette with low levels of what was regarded at the time as the major carcinogen (benzpyrene) but had not marketed it. There was concern about how to make claims for safer products that would not indict other products, and there was a keen appreciation of the difference between product innovation for public relations purposes and product innovation to reduce the hazard of the product. Finally, there was an appreciation of the spirit of sharing and comity among the British companies as regards "health-oriented" innovations in contrast to the competitive turn that had just emerged in this area in the United States.

The following exchange, during the discussion following Sir Charles's talk, illustrates the tension that knowledgeable spokesmen for tobacco companies were under at the time, because they were denying the scientific evidence that smoking is dangerous while, at the same time, they knew that the evidence was sound and that internal studies such as mouse skin-painting experiments were consistent with the publicly available results.

Mr. Reid suggested that no industry was going to accept that its product was toxic, or even believe it to be so, and naturally when the health question was first raised we had to start by denying it at the P.R. level. But by continuing that policy we had got ourselves into a corner and left no room to manoeuvre[.] [I]n other words if we did get a break through and were able to improve our product we should have to about-face, and this was practically impossible at the P.R. level. If we could ease the approach a bit, then when we did make positive contributions we could at least say so without having to crawl behind the door .

Mr. McCormickdid not quite agree that we in this country had got ourselves into that position, although it might be true of other countries [presumably, the United States]. We had more room to manoeuvre because, whatever we had said initially, in the last year or two we had been prepared to admit that there was a working hypothesis which ought to be examined. The fact that we had started with a donation to the MRC [Medical Research Coun-

cil] had indicated that we thought there might be something in it. But it was very difficult when you were asked, as Chairman of a Tobacco Company, to discuss the health question on television. You had not only your own business to consider but the employees throughout the industry, retailers, consumers, farmers growing the leaf, and so on, and you were in much too responsible a position to get up and say: "I accept that the product which we and all our competitors are putting on the market gives you lung cancer", whatever you might think privately [emphasis added]. {1102.01, p. 45}

The Phenol Crisis

On May 22, 1962, the P. Lorillard Company precipitated a crisis within the cigarette industry: it announced that it had introduced a modification to the Micronite filter of its Kent brand that selectively removed 90 percent of the phenols from mainstream smoke (8). (The filter Lorillard had announced for Kent had been produced in collaboration with Dr. Ernst Wynder of the Sloan-Kettering Institute for Cancer Research.) The fallout from this announcement was still being felt in July, and the phenol problem was a major topic of discussion at BAT's 1962 research conference in Southampton.

The phenols were regarded as irritants and probable cocarcinogens. The irritating effects of phenols were thought to account for paralysis of normal ciliary function in the tracheo-bronchial tree that cigarette smoke causes. The cilia, little hairs that line the windpipe, help cleanse the lung by moving mucus up to the throat. As reported in the Wall Street Journal , Lorillard's public relations firm issued a statement claiming that "studies on the cilia in frogs show that the workings of the cilia are unaffected by smoke from Kents with the new filter. ... Smoke from another brand of filter cigarette, it claims, slowed the cilia down by 67 percent, while smoke from an unfiltered cigaret slowed action of the cilia by 85 percent. There is no mention made [by Lorillard's public relations firm] of any research or theories on the possibility that phenols may be cocarcinogens. The statement also claims that in smoke from an unfiltered cigarette there are about 105 micrograms of phenol. By comparison, it claims, there are 35 micrograms in a filtered cigarette, 24 micrograms in an old-style Kent and 12 micrograms in Kents with the new filter" (9). Lorillard's announcement, then, meant that it sought a competitive advantage from a product that seemed to offer a genuinely "health-oriented" benefit.

Ernst Wynder had made numerous contributions to the scientific understanding of cigarettes and disease in the 1950s and 1960s, initially at

Washington University and then at the Sloan-Kettering Institute. He participated in major epidemiological and laboratory studies of smoking and health. In the mid-1950s he developed the mouse skin-painting model that was to become the cigarette industry's standard for assessing the carcinogenicity of cigarette smoke condensate.

In testimony before Congress in 1957, Wynder had pointed out that the cigarette filters of the day were not selective. That is, they did not remove toxins to any greater degree than they removed harmless smoke components (3). He also criticized some cigarette manufacturers for blending the tobaccos in their filter cigarette brands so that they produced smoke with higher levels of tar than came from the nonfilter brands of the same manufacturer. The year before Lorillard's announcement, a paper Wynder and Hoffman had published in the journal Cancer had drawn attention to the fact that the phenolic fraction of cigarette smoke had cocarcinogenic activity (10). This finding provided a rationale for seeking the selective removal of phenols.

The problems that the new Kent filter created for BAT—and, particularly, B&W—are reflected throughout the general proceedings of the Southampton conference {1102.01}. The matter was of such urgency and importance that a special session at the conference was devoted to the problem, and a full transcript of this session, including both the presentations and the discussion, was made {1102.02}. The mood at this session suggested that BAT was about two years behind Lorillard in the development of selective filtration. In his opening remarks at this session, Sir Charles Ellis acknowledged the profound embarrassment the BAT scientists had felt as well as the challenge that lay before them.

Gentlemen, we must all recognise that it is natural for our Boards to wonder why scientists of another Company should have brought to their Board's attention the possible importance of the problem of the selective filtration of phenols and that we ourselves unfortunately did not. I suspect that this has already been mentioned in their discussions and they are now awaiting anxiously for the next surprise to burst upon them. It is precisely to avoid such unpleasant situations that they provide themselves with expensive research establishments. Let us admit that it would be both natural and reasonable for our Boards to think in this way and irrespective of whether they do or do not, as may be the case, there is definitely something for us to answer. I personally feel this responsibility very keenly for I am in as good a position, if not better, than anyone to know the facts and to guess trends. If there were to be any kicks to be distributed I think they should first land on me. I suggest that this morning anyone should feel free to job [sic ] backwards and say what in their opinion should have been done. If this involves the suggestion that someone should have done some-

thing different or acted more quickly then let it be said openly and frankly—straight speaking—we are all friends.

Now the reason for raising this point in rather a direct manner is because of the overriding importance of making our Group Research a successful reality—a reality which can grapple with situations like this. Obviously, it can be successful Group Research if it expresses joint opinions of you all acting as equals with equal responsibilities. I hope our discussions this morning will help that onwards. {1102.02, pp. 1–2}

These opening remarks were followed by presentations by H. D. Anderson from Millbank, who gave an overview of the situation; by T. M. Wade, Jr., who presented the problem from the B&W perspective; and by Mr. Laporte, who commented on the problem from a Canadian point of view. In his presentation Anderson emphasized the political importance of Lorillard's going public with the connection between phenol and ciliatoxicity, since the impairment of ciliary activity was, in his view, "implicated in the cancer problem" {1102.02, p. 8}.

How important is phenol? There are two sides to this—the scientific side and the political attitude. Let's be quite certain about this: the political implications have at the moment by far the greater importance and they have been summed up very adequately in a letter from Add [Addison] Yeaman, the lawyer at B. & W. Now that Lorillards have come right out into the open and made it publicly known that the [sic ] connection between phenol and cilia activity, and because cilia activity goes back a long way and has been implicated in the cancer problem, the argument would run like this—I quote from Yeaman's letter:

"Query: In this state of knowledge is it negligence on the part of the cigarette manufacturer either

Yeaman was suggesting that B&W had legal vulnerability if it did not act to remove phenols or warn smokers of the problem. Of course, these discussions took place before the advent of warning labels, which, ironically, have had the effect of largely protecting the industry from liability under a failure-to-warn theory.

The phenol crisis presented a real-world example of the competitive problems that would arise if a tobacco company were to develop a "safe" cigarette—most notably, that all other cigarettes would be admittedly more dangerous. The potential liability issues were formidable (see chapter 7).

In his presentation Tom Wade, of B&W research and development, complained about the absence of a cooperative approach in the United States.

This is a typical example of too little, too late. This situation of ours [B&W's] is somewhat different from yours [BAT's] over here as it is now proven that certain elements of the industry will go it alone, rather than through T.I.R.C, in spite of the fact that they are members of the T.I.R.C. That leaves us in a very awkward situation in the States. Of course, we don't know the level of activity of phenols or the importance of them, but the Reader's Digest can make a powerfully good case of this; in all practical purposes they are anti-tobacco to begin with—thanks to their association with our boyfriend (Wynder). Secondly, it increases the circulation [of the Reader's Digest ] by having a good scare ..., and on top of that, they are very much insulted at the United States Government and the Federal Trades [sic ] Commission in particular for having stopped all kinds of reference to filtration in terms of total tar; we can talk about the foolish side if we want to, taste and flavour and so on, but the implications from the medical standpoint are being carried by the Reader's Digest in opposition to the Federal Trades [sic ] Commission. ...

We have gone through to the battle of the total tar and we are now in the battle of the phenols—tomorrow it will be something else. Of course Wynder does not believe that a cigarette will ever be completely safe. {1102.02, pp. 15–16}

The Reader's Digest had a long history of publishing articles about smoking and health, and since the mid-1950s it had published important articles about the relative tar and nicotine deliveries of various filter and nonfilter cigarettes. In 1955 the Federal Trade Commission put a stop to the most obvious health claims in cigarette advertising. References to "health protection" and to "reduced irritation" disappeared and were replaced by what Wade calls "the foolish side," an emphasis on taste and flavor. Wade's complaint, in part, was that although the FTC had barred the manufacturers from making the most flagrant health claims in their cigarette advertising, the Reader's Digest was still publishing articles on the relative benefits of various cigarette brands in terms of tar delivery (11). These articles were, inevitably, detrimental to some particular companies' brands, and the companies could not respond in their advertising.

The phenol crisis illustrates the complex environment in which the R&D scientists operated. Most important, it illustrates the fact that the tobacco companies' scientific agenda was set by competitive and public relations needs as much as by scientific priorities.