Tobacco Documents Online

xii TABLE OF I~E(vTS Reduction of Tar and Nicotine Levels of Austrian Cigarettes During the Last 15 Years ................................. .89 H. Kuhn, and H. Kius Cigarette Smoke and Tar and Nicotine Levels in Sweden During the Decade 1964-1974 ................................ L. Hjcrii Recent Stw;ics on Less Harmful Cigarettes at the University of Waterloo, Ontario, Canada .............................. 10 W. H. Cherry Less Harmful Ways of Smoking ............................. 111 Irwin D. J. Bross PART 11: RELATION OF SMOKE COMPONENTS TO CANCER Methodology for Determining Beneficial Effect of Less Harmful Cigarettes on Lung Cancer Risk ........................... 119 Lawrence Garfinkel Chemical Studies on Tobacco Smoke. XXXIX. On the Identification of Carcinogens, Promoters and Cocarcinogens . . . . . . . . . . . . . . . . . . 125 Dietrich Hoffmann, Irwin SchmcltZ, Stephen S. Hecht, and Ernst L. Wynder Chemical Identification in Tobacco Smoke of Carcinogens, Promotcrs anc: Cocarcinogens in Tobacco Smoke ................ 147 H: P. ; larkc Discussio<<- Identification of Carcinogens, Tumor Promoters and Cnc.i:cinogens in Tobacco Smoke .......................... 155 ,M.R. Guerin Chcm;cal and Biological Identitication of Tumorigenic Components ofTubacco ......................................... 161 Fred G. Bock, and T.C. Tso Tumor r'romoting Activity of Agriculture Chemicals ............... 175 Fred G. Bock, and T.C. Tso Ch.-nicai Studies on T obacco Smoke. XL. Identification of Carcinogens : n Tob=cco ................. ....................... 191 Stca hen S. f!c:cht, Lrwir, Schn.c{tt, Dictrich Hoffmann, and Enut L. Wynder N ~- - ~_ _ -- ---- -- -- ~3 - . G~ C17 ~ -t~ CT+

TABLE OF CONTENTS xtu Discussion of Chemical and Biological Identification of Tumorigenic Components of Tobacco ........................ 203 D. Bhaskara Reddy Biological Evaluation of Carcinogens in Tobacco Smoke .............. 204 W. Dontcnwill The Biological Evaluation of Substitute Smoking Materials as Exemplified by 'NSM................................... 221 D.M. Conning Cocarcinogenic and Tumor-Promoting Agents in Tobacco Carcinogcnesis (Abstract) ................................ 229 B.L. Van Duurcn PART Iil: RELATION OF SMOKE COMPONENTS TO CARDIOVASCULAR DISEASE Introduction to Smoking and Cardiovascular Disease ................ 231 Wilbert S. Aronow :titethodology for Determining the Possible Beneficial Effect of a Less Harmful Cigarette on the Risk of Cardiovascular Disease ........ 237 Tavia Gordon Relative Risks of Myocardial Infarction, Cardiovascular Disease and Peripheraf Vascular Disease by Type of Smoking .............. 243 Manning Feinlcib, and Roger R. Williams Epidemiologic Studies on Smoking in Cerebral and Peripheral Vascular D isease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 William B. Kannel Epidemiologic Studies in the United Kingdom on Smoking and Coronary Heart Disease ................................. 275 D.D. Reid Smoking and Peripheral Arterial Disease : ....................... 281 Andreas Koch The Interrelationship of Tobacco Smoke Components to Hyperlipidemia and Other Risk Factors . . . . . . . . . . . . . . . . . . . . . . . 285 Thomas R. Dawber

xiv TABLE ONTENTS Discussion of Dr. Dawber's Paper (The Interrelationship of Tobacc~ Smoke Components to Hyperlipidemia and Other Risk Factors). ... 293 Oglesby Paul The Evidence for Nicotine as an Etiological Factor in Cardiovascular Disease .............................. .297 'Helmut Schievelbein Discussion of Dr. Schievelbein's Paper (The Evidence for Nicotine as an Etiological Factor in Cardiovascular Disease) ............... 19 Henry C. McGill, Jr. Nicotine: An Etiological Factor for Coronary Heart Disease ........... 3 Peter Hill Carbon Monoxide and Cardiovascular Disease ..................... 321 Wilbert S. Aronow Discussion of Dr. Aronow's Paper (Carbon Monoxide and Cardiovascular Disease) .................................. 329 Poul Astrup Studies on Carbon Monoxide and Nicotine........................ 331 Poul Astrup Monitoring of Carboxyhemaglobin in a Cardiovascular Clinic........... 343 R.W. Sillett, J.A. McM. Turner, and K.P. Ball Carbon Monoxide as an Etiological Agent in Arterial Discasc-Some Human Evidence ............................ 349 Nicholas J. Wald Evidence for Components Other Than Carbon Monoxide and Nicotine as Etiological Factors in Cardiovascular Disease............ 363 Gardner C. McMillan Significance of Nicotine, Carbon Monoxide and Other Smoke Components in the Development of Cardiovascular Disease .......... 369 William B. Kannel, and William P. Castelli

TABLE OF CONTENTS xv PART IV: RELATION OF SMOKE COMPONENTS TO CHRONIC PULMONARY DISEASE Long-Term Surveillance of Tobacco Substitutes ................... 383 Walter W. Holland Relative Risks of Various Tobacco Usages for Emphysema, and/or Chronic Bronchitis ..................................... 389 I.T.T. Higgins Types of Tobacco Usaae and Chronic Respiratory D:sease ............. 407 T. Abelin Chronic Disease-The Etiological Factors in Smoke ................. 415 Tore Dalhamn Chronic Lung Disease and Tobacco Smoke Exposure-Concepts on Etiological Factors ..................................... 427 Ragnar Rylander Bioassaying Candidate Less-Hazardous Cigarettes: Methods for Early Detection of Peripheral Airway Dysfunction .................... 433 Louis Diamond PART V: TECHNOLOGICAL ASPECTS OF REDUCING SPECIFIC COMPONENTS IN CIGARETTE SMOKE, AND CONSUMER ACCEPTANCE ~ OF SUCH REDUCTION Chemical Studies on Tobacco Smoke. XXXVII. Determination of Tar, Nicotine, and Carbon Monoxide in Cigarette Smoke. A Comparison of International Smoking Conditions ......................... 441 Klaus D. Brunnemann, Dictrich Hoffmann, Ernst L. Wynder, and G.B. Gori Approaches to the Reduction of Total Particulate Matter (TPM) in Cigarette Smoke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... . 451 Gio B. Gori Possibilities for the Reduction of Nicotine in Cigarette Smoke ..........463 H. Kuhn and H. Klus Selective Reduction of Tumorigenicity of Tobacco Smoke. III. The Reduction of Polynuclear Aromatic Hydrocarbons in Cigarette Smoke ...................................... 495 Dictrich Hoffmann, and Ernst L. Wynder i i

xvi TABLE OF qTENTS Discussion: Complexing Compounds Caffeine - PAH ............ . .505 J¢rgen Lam 1 Vapor Phase Smoke Modification-An Under-Utilized Technology .... 507 Donald Tigglebeck Technological Aspects as to Our Ability Toward, and Consumer Acceptance of, Reducing Phenolic Compounds in Cigarette Smoke (Abstract) ........................................... Charles J. Kensler , Cilia Toxic Components in Cigarette Smoke ...................... 51. Sam P. Battista Selective Reduction of Tumorigenicity of Tobacco Smoke. IV. Approaches to the Reduction of Nitrosamines and Aromatic Amines ............ 535 Stephen S. Hecht, T.C. Tso, and Dietrich Hoffmann PART VI: SUMMARY OF WORKSHOP I( RECOMMENDATIONS. ..547 Ernst L. Wynder INDEX ........................................... 555

NOTICE •aY be This materizl ~ prote i ,d by U.S. law (Tit}e r7 THE EPIDEMIOLOGY OF THE LESS HARMFUL CIGARETTE Ernst L. Wynder, Margaret Mushinski, and Steven Steliman Retrospective data on patients with tobacco-related diseases and controls in- dicate a reduced risk for lung, oral cavity and larynx cancers in male, long-term filter smokers and for lung and larynx cancer in long-term filter smoking females. This reduction is due in part to lower "tar" yield of filter cigarettes (the "less harmful cigarette"-LHC) and in part to the increasing number of ex-smokers. Nicotine is implicated in the etiology of peripheral vascular disease and more re- search is needed in the epidemiology of myocardial infarction and emphysema in relation to smoking. Mortality data from England and the U.S. also support the reduced risk for lung cancer associated with smoking the Li-tC and the data indicate a decreasing death rate in the younger cohorts of men. Further refinement of the LHC and its effects are of obvious importance and continued attention should be paid to the effects of these cigarettes on chronic pulmonary disease, emphysema, and heart disease. The tobacco and health problem continues and new approaches are obviously necessary in order to cornbat this public health dilemma. A multi- phased approach is suggested which would include increased anti-smoking educa- tion, increased and more prominent smoking cessation programs and clinics in hospitals and other medical facilities and the further development of a truly less hazardous tobacco product. INTRODUCTION From the very first, epidemiological studies on tobacco-related lung cancer clearly demonstrated that this disease was quite rare among nonsmokers and that its incidence directly increased as the individual's consumption increased (1-3). Similar correlations have been shown for other tobacco-related cancers (4). The established dose-response relationship of these cancers to tobacco smoke exposure has provided the basis for the likely assumption that as smoke exposure decreases, either by smoking less or by smoking a cigarette lower in smoke yield, the risk of developing tobacco-related cancers decreases proportionately. An obvious problem in measuring or predicting the risk of developing cancer is the long latency period between first exposure to a carcinogen and clinical manifestation. Often there is an interval of many years. For example, a long term heavy smoker who gives up smoking has an increased risk of lung cancer for at least ten years over that of a person who has never smoked (5). The average age of a male lung cancer patient today is approximately 60 years old. Since this per- son has been smoking for about 40 years, his smoking history includes a com- bination of the old high "tar" cigarettes and the newer low "tar" cigarettes, with the latter type having been smoked for considerably less time than the former. For this group, therefore, the influence of or relationship between smoking only

2 WYNDER, MUSHINSKI AND -TELLMAN the low "tar" cigarette and the development of lung cancer cannot precisely measured. One must take into account the number of years and inten. v of high "tar" cigarette smoking and weigh this in relation to the length of tin the per- son has smoked the lower "tar" yield cigarettes. Such a formula woulc 'low in- vestigators to determine an exposure level for each individual based or. aecific smoking histories. The effect of this latency period is likely to differ for cardiovascular ~•ents because, for these diseases, at least a portion of the effect of smoking appt. s to be an acute one. The effect may also be different for emphysema, a conc on which seems irreversible, once a certain stage of pathology has been reachec t the present time, no data are available which have measured the effect of sm• ing the less harmful cigarette (LHC) on the development of emphysema. A fc studies have been conducted illustrating this effect on cardiovascular disease more specifically, peripheral vascular disease (PVD). The data available on tobacco- related cancers is more extensive, however. In respect to tobacco carcinogenesis, two major events have taken place re- garding the "less harmful cigarette": a) smoke condensate yields of American cigarettes and, indeed of cigarettes throughout most of the world, have been re- duced during the past 25 years and, b) the carcinogenic activity of "tar," both from filter and nonfilter cigarettes, as measured on mouse skin and reflected by its chemical constituents, is lower today than it was 25 years ago (6, 7). On this basis, one may predict two epidemiological developments: Firstly, the risk of de- veloping cancer for low "tar" (largely filter) cigarette smokers (at least for those who have smoked filter cigarettes for ten years or more) will be lower than for those who have continued to smoke high "tar" (largely nonfilter) cigarettes. Such a reduction would be due to the fact that the "tar" yield of filter cigarettes is generally lower than that of nonfilter cigarettes ( Fig.1). The carcinogenic activity of such cigarettes, however, is not different when measured on a gram for gram basis on mouse skin (8). It should be explained, however, that even those who now smoke nonfilter cigarettes should have a lower risk than those who smoked nonfilter cigarettes 25 years ago because these cigarettes also have a reduced "tar" yield than previously was the case. Secondly, the risk will be lowest for those in the age groups in which the LHC has been smoked for the longest period of time. Thus, when we compare the death rate from lung cancer in the early 1970's with that of the early 1960's, we should observe a reduction or a slower increase among the younger male population and a continued increase among the older male population. Such a change will be less noticeable for women because even though they have smoked the less harmful cigarette relatively more often and for a long- er period than men, there are significantly more women smokers in even the younger cancer affected age group now than there were 15 years ago (9). With these basic assumptions, let us determine whether they can be substan- tiated by available data. On the basis of ongoing studies being conducted in eight cities in the United States, we find that the rates of lung cancer among long-term filter cigarette smokers are somewhat lower for all cancer sites for which we have obtained sufficient data to permit statistical analysis. This includes cancers of the lung, oral cavity, and larynx (Figs. 2-4). We have been able to demonstrate a

EPIDEMIOLOGY OF LESS HARMFUL CIGARETTE 3 FIG. 1. Filter and nonfilter tar yields in U.S. cigarettes, 1958-1973. FIG. 2. Relative risk of lung cancer (Kreyberg 1) for current filter and nonfilter cigarette smokers-men-New York, Los Angeles, and Houston, 1970-1973. wv.n.rv., ro~cro.+o+a~o-tn FIG. 3. Relative risk of oral cavity cancer for male cigarette smokers by number and type of cigarettes smoked daily.

4 4Nr. Iv75 FIG. 4. Relative risk of laryngeal cancer for male cigarette smokers by number and type of - cigarettes smoked daily. , similar reduction in risk for lung as well as for larynx cancer among women long- term smokers of filter cigarettes (Fig. 5) (9). These data, similar to those reported by Bross and Gibson (10), show that filter cigarette smokers of ten years or more have reduced risk of developing lung, larynx and mouth cancer compared to those who have continuously smoked nonfilter cigarettes.This finding can be attributed at least in part to the lower'rtar" yield of filter cigarettes. An additional benefi- cial influence must be given to the finding of an increasing number of ex-smokers. Obviously, the earlier in life smoking cessation takes place, the sooner the can- cer risk is reduced. Thus, more attention should be paid to strengthening the smoking cessation message as well as making such clinics more widely available. .e ~ ®.or rwa. Lq ~~.=0 Z~.Sp .~uyr K+ a.. arettes smoked daily. WYNDER, MUSHINSKI AND FTELLMAN 'Risl reYtM to 1.0 fer ransTae.n ]I. ,s. ,eur..t *o Lo ro+.~ ~ortn FIG. 5. Relative risk of lung cancer for female cigarette smokers by number and type of cig- We are, of course, also interested in determining the risk for individuals who smoke cigarettes of less than 16 mg'rtar." Clearly, there are differences in "tar"

EPIDEMIOLOGY OF LESS HARMFUL CIGARETTE 5 yield among filter cigarettes. In general, the "tar" yield of filter cigarettes has de- creased during the past 15 years. Cigarettes with 16 mg or less of tar, however, have not been smoked by a sufficient number of people for a long enough period of time to give us reliable information on the risks associated with these cigarettes alone (Fig. 6). FIG. 6. Brands delivering 16 mg tar or less-market share vs. year. At present there are no similar data for emphysema nor for myocardial in- farction. Data from the National Heart and Lung Institute suggest that the death rate from myocardial infarction is decreasing (11). If this is the case, it remains to be shown whether this is due to improved treatment of hypertension, reduction of serum cholesterol levels, more ex-smokers among adults now than in the past, or a reduction in the nicotine levels of cigarettes. This result cannot be due to a reduction in carbon monoxide in cigarette smoke because this component has not been reduced in the last 25 years, nor has it significantly differed for filter and nonfilter cigarette smoke. Obviously, the epidemiology of myocardial infarc- tion is relatively complicated because it is related to several known risk factors. We have, however, obtained data on peripheral vascular disease (Fig. 7). These data suggest that the risk of developing PVD is lower among filter cigarette smokers than for nonfilter cigarette smokers-a finding which suggests that nico- tine is indicated in the etiology of this disease (12). Koch has obtained similar 1 data in an independent study in Germany (13). EFFECT ON DEATH RATES The risk• associated with the LHC can also be measured by comparing'death rates from tobacco-related diseases in the 1970's to those of the 1950's and 1960's. According to previous assumptions, such a reduction should be demonstrated particularly for the younger male population. Additionally, a greater reduction should be indicated for men than for women since, as we have previously stated, what has been gained by more women than men smoking the LHC has been lost