92% TaBLE A15.—Studies concerning the relationship of smoking to infectious respiratory disease in humans (Actual number of cases shown in parentheses) SM = Smokers NS = Nonsmokers Author, year, Number and Data country, type of collection Results Comments reference population Mills, 118 male and Hospital Cases Controls The author stated that 1950, female patients Interview. Mean age 49.6 49.6 there was a U.S.A. with pneumonia NS ccc cece cece ese eae 15.25 25.21 significant difference (167). and 472 healthy Cigarettes only 63.56 62.33 in tobacco usage individuals from Mixed 21.19 22.46 between the “random” sample. two groups. Lowe, 520 male and Interview by Males Females Cigarette smokers 1956, 185 female trained Cases Controls Cases Controls include pipe smokers. England tuberculosis social NS cece cece renee eee eee 2.5 8.1 37.8 61.4 The author noted a (157). patients and 419 worker. Cigarettes/day: 1-9 ...... 9.2 12.9 20.5 25.7 significant deficiency male and 249 W019 eee eee 38.1 35.6 30.8 20.5 of non- and light female control 20-29 oo eee eee eee 29.4 27.4 smokers and an outpatients. 80-39 cee eee eee eee 11.3 9.3 11.4 2.4 excess of heavy oo | 9.4 6.7 smokers among the cases Dowling, Individuals Interview and Exposed to placebo Exposed to infectious agent No statistically etal, exposed to medical Percent Percent significant 1957, “infectious examination. developing developing differences U.S.A. cold agent” Number “cold” Number “cold” noted. (72). and placebo. NS wee ccc eee cee eee 111 10 328 34 SM woe ccc ee eee eves 18 14 249 35 LEZ TaBLE A15.—Studies concerning the relationship of smoking to infectious respiratory disease in humans (cont.) (Actual number of cases shown in parentheses) SM = Smokers NS = Nonsmokers Author, year, Number and Data country, type of collection Results Comments reference population Boake, Parents of Interview Number of No statistically 1958, 59 families. Person- reapiratory Illnesses/ significant U.S.A. years ilineasea person-years differences (38). NS cece eee eee enes (24) 120 624 5.2 noted. Cigarettes/day: 1-10 ......... (19) 99 529 5.3 U2-20 Lecce eee eee eee (25) 108 486 4.5 S20 cece eee cnn eee (19) 99 424 4.3 Pipe, cigar ...-.....ees renee (14) 12 304 4.2 Shah Tuberculosis Survey, X-ray, Tuberculous Normal or + Numbers in et al, institute and by X-ray . nontuberculous parentheses 1959, employees. interview. NS coe e cece ccc ee eee ee t10 (19.7) 178 (168.3) represent figures India SM owe ee eee eee 86 (26.3) 215 (224.7) “expected” by use of (205). 2x2 contingency table. Tuberculous employees were found to have significantly fewer nonsmokers and more smokers. 8zz TABLE A15.—Studies concerning the relationship of smoking to infectious respiratory disease in humans (cont.) (Actual number of cases shown in parentheses) SM = Smokers NS = Nonsmokers Author, year, Number and country, type of Data reference population collection Results Comments Brown 306 male and Interview Smoking habits prior to diagnosis Data presented only et al., female Tuberculous patients Controls on Queensland 1961, tuberculosis (percent) (percent) sample. Australia elinie NS coc cee eee cee nee . 9.1 19.9 The authors noted (4). patients, Cigarettes/day: 10.5 15.4 that the 221 male and 10-19 84.3 19.5 significant difference female 20-29 26.3 25.8 between the outpatients. 30-39 1.2 5.4 patients and >40 Lo. 6.2 9.1 controls was not Pipes 5.9 4.6 present when the groups were matched for alcohol intake. Haynes 191 male Interview Average number of respiratory illnesses/10 students etal, prep school (adjuated for age) 1966, students. All severe lower U.S.A. All All severe or combined (108). respiratory respiratory respiratory episodes episodes episodes NS (99) ci cec cece cence eee 11,1 1.6 0.36 SM (92) cee eee eee eee 20.2 6.7 3.34 Parnell 47 smoking- Interview Median number of illnesses/student The authors noted et al., nonsmoker pairs and health All All that these 1966 of student nurses service respiratory other differences were Canada matched for age records. diseasest illnesses statistically (181). and parents’ NS (47) 2.08 2.99 significant. occupational SM (47) 2.54 5.00 ¢ Particularly class. tracheitis, bronchitis, and pneumonia. 672 Tapie Al5.—Studies concerning the relationship of smoking to infectious respiratory disease in humans (cont.) (Actual number of cases shown in parentheses) SM = Smokers NS = Nonsmokers Author, year, Number and Data country, type of collection Results Comments reference population Peters 1,496 Harvard Medical history, Number of visits to student health unit for respiratory illness/ student t p<0.001. et al., and chart review, (common. colds, pharyngitis, bronchitis, laryngitis, 1967, 370 Radcliffe and pneumonia—not allergic rhinitis) U.S.A. students. questionnaire. Harvard Radcliffe (183). DS 1.44 (771) 1.44 (193) SM nce eee eee eee $2.27 (725) 2.27 (177) <2 years smoked .......- 2.00 BHA eee tenes 2.30 DB eee eee eee 2.50 Finklea 1,811 male Questionnaire Heavy smokers—21 percent more clinical illnesses than nonsmokers; The authors also et al., college prior to 20 percent more requiring bed rest than nonsmokers noted that: 1969 students. A,/HK/68 ” Light smokers—10 percent more clinical illnesses than nonsmokers; (a) Smokers U.S.A. epidemic and 7 percent more requiring bed rest than nonsmokers. exhibited (83). follow-up on serologic morbidity. evidence of increased subclinical A,/HK/68 infection. (b) There was no difference in the vaccination status between smokers and nonsmokers. TABLE A16.—Complications developing in the postoperative period in patients undergoing abdominal operations Men over 20 Percent Percent broncho- Percent Group Cases chest Percent pneumonia total clear bronchitis and complication atelectasis rate Smokers .............-.+.-.- 800 41,7 53.0 5.3 58.3 Light Smokers .............. 180 68.4 27.7 3.9 31.6 Nonsmokers ................ 66 92.5 6.0 1.5 1.5 Women over 20 Smokers ............0 0005: 23 89.1 43.5 17.4 60.9 Light Smokers ............. 62 WS 20.9 1.6 22.5 Nonsmokers ..........-.00005 518 88.8 8.1 3.1 112 Source: Morton, H. J. V. (173) TABLE A17.—Arterial oxygen saturation before and after operation Arterial oxygen saturation (percentage) Case Before Group number operation Day 1 Day 2 Day 3 1 94 93 94 2 94 93 94 Nonsmokers 2... 0.0.00 cece eee eee 3 96 93 94 4 95 90 94 5 94 90 93 6 95 91 89 91 7 92 89 81 89 Smokers . 1.2.00... 0. cee eee eee 8 91 89 85 89 9 93 91 88 92 10 30 87 88 92 Source: Morton, A. (172). 230 CHAPTER 4 Cancer Contents Introduction ..... 0.0 ce ce eee eee teens Lung Cancer Nickel Chromium Pathological Studies Tracheobronchial Implantation and Instillation ........... 0.0 c eevee Inhalation .. 0.0... cee eee Reduction in Tumorigenicity .............. Summary and Conclusions ....... 00.052. e eee eens Cancer of the Larynx Pathological Study Oral Cancer Epidemiological Studies ... 0.0.0... c eee ee ee ee eee Prospective Studies 00... 0... cee eee ee eee eee Retrospective Studies ........ 0... cee eee ees Lung Cancer Trends in Other Countries ............ Histology of Lung Tumors .......... 00.002 eee eee ee Lung Cancer Relationships in Women ............-. Lung Cancer, the Urban Factor, and Air Pollution.... Lung Cancer and Occupational Hazards ............ Uranium Mining ........... 2.00 ee ee Other Occupations... 0.0.0... 02. AsbestoS. 2... ccc eee eee AYSCNIC. 2 eee enn eee Pulmonary Carcinogenesis ....... 0... ee ee eee General Aspects of Carcinogenesis ............. Polynuclear Aromatic Hydrocarbons ....... Nitrosamine Compounds................5- Pesticides and Fungicides................-. Radioactive Isotopes ...........0000e sere Inhibitors of Ciliary Movement ........... Experimental Studies ...............0 eee eee Skin Painting and Subcutaneous Injection... Tissue and Organ Culture .............4. Epidemiological Studies ...... 0.0... ec eee eee eee Experimental Study .. 02.6... 0c cece eee ees Summary and Conclusions .......... 2. eee eee eee Epidemiological Studies ...... 0.0... eee eee ee ee eee Experimental Studies ......... 00.50 e eee eee eee eee Summary and Conclusions ..... 6.0.52. eee eee eee Cancer of the Esophagus ................000ccceeeeeue Epidemiological Studies ...................00 0000, Pathological Study ........0. 000... cee cee cee eeee Experimental Studies ....... 0000.0 0c. e cece ce eee Summary and-Conclusions .............0....00000, Cancer of the Urinary Bladder and Kidney .............. Epidemiological Studies (Bladder) ................. Epidemiological Studies (Kidney) ................. Experimental Studies ........00.0. 0.0... 0c cece eee Summary and Conclusions .........0..... 00000000, Cancer of the Pancreas ..... 0.00. cece cece cee ccceee Summary and Conclusions .............. 000.00 ceee References ......... 00.0 ccc eve ec cece ev cece ee ncnnnces FIGURES 1. Lung cancer, Finland and Norway ................... 2. Percent of smoking dogs with tumors ................ 3. Percent of lung lobes with tumors in smoking dogs ..... 4. Effects of chronic cigarette smoke inhalation on the A3. Ad, 10. 234 hamster larynx ....... 0.0000. cc cee cece ce eeeee LIST OF TABLES Lung cancer mortality ratios ................004, Lung cancer mortality ratios for males by duration of cigarette smoking .............. 0.00.00 ceeeee Outline of methods used in retrospective studies of smoking in relation to lung cancer ............. Group characteristics in retrospective studies on lung cancer and tobacco use ...............00.00 ce Annual means of total lung cancer mortality and sex ratios for selected periods in Finland and Norway Epidemiologic and pathologic investigations concern- ing smoking and histology of lung cancer ....... Grouping of pulmonary carcinomas .............. Tumor prevalence among males and females 35-69 years of age, by type of tumor and smoking CaleQOry . 6. ke cee ce cece nuevas Epidemiologic investigations concerning the relation- ship of lung cancer to smoking, air pollution, and urban or rural residence ...................04, Pathologic and cytologic findings in the tracheo- bronchial tree of smokers and nonsmokers ...... Page 289 289 292 292 293 293 293 296 296 299 299 299 299 245 274 274 284 241 244 323 329 246 247 334 250 2538 259 Al4, Al5, Al6. 17. 18. 19, 20. A21, A22, A23, A24, 20. 26. 27, A28. A28a, LIST OF TABLES (Continued) (A indicates tables located in appendix at end of chapter) Identified or suspected tumorigenetic agents in cigarette smoke .........0. 0. ccc eee eee eee Autopsy studies concerning the presence of radio- activity in the lungs of smokers .:............. Experiments concerning the effects of the skin paint- ing or subcutaneous injection of cigarette smoke condensate or its constituents upon animals ..... Experiments concerning the effect of cigarette smoke or its constituents on tissue and organ cultures .. Experiments concerning the effect of the instillation or implantation of cigarette smoke or its constitu- ents into the tracheobronchial tree of animals .... Experiments concerning the effect of the inhalation of cigarette smoke or its constituents upon the respiratory tract of animals .................. Data on pedigreed male beagle dogs of groups F, L, H, hand N .... ec eens Summary of principal cause of death (days No. 57 through No. 875) in dogs of groups F, L, H, h and N Data on dogs with lung tumors indicating type of tumor and lobe in which the tumor was found .... Laryngeal cancer mortality ratios — prospective studies 2.0... ccc ene ee eens Outline of retrospective studies of tobacco use and cancer of the larynx .....0...... 0.0 cece cece eee Summary of results of retrospective studies of tobacco use and cancer of the larynx ...............005- Number and percent distribution by relative fre- quency of atypical] nuclei among true vocal cord cells, of men classified by smoking category ..... Number and percent distribution, by highest num- ber of cell rows in the basal layer of the true vocal cord, of men classified by smoking category .... Deposition of }#C-labeled smoke particles in particu- lar regions of the respiratory tract ............ Classification of the five registered stages of epithe- lial changes at the larynx .............. 0.2.0.4. Oral cancer mortality ratios—prospective studies. . Outline of retrospective studies of tobacco use and cancer of the oral cavity ............. 05. eee Summary of results of retrospective studies of smok- ing by type and oral cancer of the detailed sites. . Page 265 335 337 343 346 349 270 271 272 278 354 358 359 360 282 283 286 361 368 235 LIST OF TABLES (Continued) (A indicates tables located in appendix at end of chapter) Page 429. Experimental studies concerning oral carcino- WENESIS oo eee eee 371 30. Esophageal cancer mortality ratios—prospective studies occ eee een eens 290 A381. Summary of methods used in retrospective studies of tobacco use and cancer of the esophagus ...... 375 A3la. Summary of results of retrospective studies of to- bacco use and cancer of the esophagus ......... 378 A382. Atypical nuclei in basal cells of epithelium of esoph- agus of males, by smoking habits and age ...... 379 A33. Atypical nuclei in basal cells of epithelium of esoph- agus of males, by amount of smoking and age .... 386 34, Kidney and urinary bladder cancer—prospective studies 2... ec ee eee eee 294 A35. Summary of methods used in retrospective studies of smoking and cancer of the bladder ............ 381 A35a. Summary of results of retrospective studies of smok- ing and cancer of the bladder ..........-...... 383 26. Pancreatic cancer mortality ratios—prospective StUdIeES oe eee eee ee eee 298 236 INTRODUCTION During the early vears of this century, a number of pathologists and clinicians reported a dramatic increase in the incidence of lung cancer, Autopsy studies and studies of lung cancer death rates re- vealed a significant increase beginning prior to World War I and continuing during the ensuing years. This epidemic of lung cancer continues to the present day, with nearly 60,000 deaths expected from this disease in the United States during 1970. Beginning in the 1920’s, a number of reports appeared which suggested a relationship between lung cancer and tobacco smoking (4, 203, 278). Since that time, many clinical and epidemiological studies have been published which confirm this relationship. The 1964 Report (291) contains a thorough review and analysis of the data available at that time as well as an excellent discussion of the considerations necessary for their evaluation. Major epidemiological studies have demonstrated that smokers have greatly increased risks of dying from lung cancer compared to nonsmokers. An increased risk of lung cancer has been found for every type of smoking habit investigated, but two character- istics of the risk are particularly evident: The risk is much greater for cigarette smokers than for smokers of pipes and cigars, and among cigarette smokers a dose relationship exists. That is, the more one smokes, as measured by total pack-years of smoking, present level of smoking, degree of inhalation, or age at start of smoking, the greater is the risk. It has also been shown that the risk of lung cancer among ex-smokers decreases with time almost to the level of nonsmokers; the time required is dependent on the degree of exposure prior to cessation. Pathologists have found that the squamous cell or epidermoid form of lung cancer is the most prevalent one in cigarette smoking populations and that this form accounts for a major portion of the rise in lung cancer deaths (75.4). Such studies have also indi- cated a lower prevalence among smokers for oat-cell and adeno- carcinomas of the lung than for the squamous form, but in most studies a higher frequency of these tumors is found among smokers than among nonsmokers. Smoking has been implicated in the development of other types of cancer in humans. Among these is cancer of the larynx. A num- 237 ber of epidemiological studies have demonstrated increased mor- tality rates for laryngeal cancer in smokers, particularly cigarette smokers, compared with nonsmokers. Autopsy studies have re. vealed that a clear dose-relationship exists between smoking and the development of cellular changes in the larynx, including carei- noma in situ. Cancers of the mouth and oropharynx have been found to be more common among users of all types of tobacco than among abstainers. Although smoking is a definite risk factor in the de- velopment of malignant lesions of the oral cavity and pharynx, its relative contribution in conjunction with other factors such as poor nutrition and alcohol consumption has not been fully clarified, Similarly, although smokers are more likely to develop carci. noma of the esophagus than nonsmokers, the relative additional] contribution of smoking in conjunction with nutritional factors and alcohol consumption requires clarification. Smokers have been found to be more at risk for the development of cancer of the urinary bladder than are nonsmokers, and there is evidence to suggest that some smoking-induced abnormal meta- bolic product or abnormal concentration of a metabolic product may be responsible for this increased risk. In addition, cancer of the kidney is apparently more common in smokers than in non- smokers, but the epidemiologic evidence for this relationship is not as definite as for bladder cancer. Epidemiological studies have indicated an association between smoking and cancer of the pancreas. The significance of this rela- tionship is unclear at this time. Experimental studies have demonstrated the carcinogenicity of the condensate of tobacco smoke, or “tar.” This material, when painted on the skin of animals, leads to the development of squam- ous cell tumors of the skin. Researchers have shown that this condensate contains substances known as carcinogens, capable of inducing cancers. Among these carcinogens are several chemicals which have been identified as tumor initiators, that is, compounds which initiate changes in target cells and also tumor promoters, or compounds which promote the neoplastic development of initi- ated cells. Other, as yet unidentified, factors are presumably also involved because the sum of the carcinogenic effects of the known agents does not equa] that of cigarette smoke condensate. Numerous experiments have been performed in which whole cigarette smoke, filtered smoke, or certain constituents of smoke, such as the “tar,” are administered by varying methods to animals or to tissue and cell cultures in order to investigate the neoplastic- inducing properties of cigarette smoke, Particular difficulty has been encountered in experiments which have attempted to deliver 238 whole cigarette smoke to the larynx and into the lungs of experi- mental animals. This has resulted in the use of other methods such as the implanting of pellets containing suspected carcinogens and the instilling into the trachea of suspected carcinogens as such, or adsorbed onto fine inert particulate matter as a carrier. The dif- ficulty with the inhalation studies has been twofold. First, the animals, particularly the smaller species such as the rat, frequently die from the acute toxic effects of the nicotine and carbon monoxide in the tobacco smoke. Second, the upper respiratory tract of experi- mental animals, particularly the nose, is much different from anal- ogous human structures, resulting in a more efficient filtration of smoke in the upper respiratory tract. Nevertheless, in rodents and canines, progressive changes apparently indicative of ultimate neo- plastic transformation have been identified in the respiratory tract. Recently, two studies in different species and in different target organs have been reported concerning the development of early in- vasive cancer following the prolonged inhalation of cigarette smoke. Auerbach and his coworkers (11) trained dogs to inhale cigarette smoke through a tracheostoma. After approximately 29 months of daily exposure, these investigators found a number of cancers of the lung. Dontenwill (76) in the second of these two studies, exposed ham- sters to the passive inhalation of cigarette smoke over varying and prolonged periods of time. He observed the development of pre- malignant changes and, ultimately, invasive squamous cell cancer of the larynx. LUNG CANCER Cancer of the lung in the United States accounted for 45,383 deaths among males and 9,024 deaths among females in 1967 (289). It is presently estimated that approximately 60,000 people will die of lung cancer during 1970. The alarming epidemic of lung cancer is a relatively recent phenomenon. Death rates for lung cancer (ICD Codes 162, 163) rose from 5.6 (per 100,000 resident population per year) in 1939 to 27.5 in 1967 (289, 290). This rapid increase followed the in- creased use of cigarettes among the United States population. The increase has occurred principally among males, although more re- cently females have shown a similar rising pattern. The converging evidence for the conclusion that cigarette smok- ing is the major cause of lung cancer is derived from varied types of research including epidemiological, pathological, and laboratory investigations. 239 EPIDEMIOLOGICAL STUDIES Numerous epidemiological studies, both retrospective and pros- pective, have been carried out in different parts of the world to investigate the relationship between smoking and cancer of the lung. These studies are outlined in tables 1, 2, A3, and A4. Prospective Studies The major prospective studies concerning the relationship of smoking and lung cancer are presented in table 1. In all, these investigations have studied more than a million persons from a number of different populations for up to 10 years. These studies show increased lung cancer mortality ratios for cigarette smokers of all amounts ranging from 7.61 to 14.20 among male smokers as compared to nonsmoking males. The one major prospective study of female cigarette smokers reveals an overall mortality ratio of 2.20 (118). Also uniformly present in these studies is a dose-related increase in the mortality from lung cancer with increasing amounts of cigar- ettes smoked per day. Other measures of exposure show similar trends. Hammond (118) reported increased mortality ratios asso- ciated with increased inhalation (table 1) as well as with increased duration of smoking (table 2). .. Ex-smokers show significantly lower lung cancer death rates than continuing smokers. In their study of more than 40,000 British physicians, Doll and Hill (74, 75) noted a decrease in lung cancer mortality rates with increasing time since smoking stopped (table 1). During the past 20 years, half of all the physicians in Britain who used to smoke cigarettes have stopped smoking. While the death rates from lung cancer rose by 7 percent among all men from England and Wales during the period from 1953-57 through 1961- 65, the rates for male doctors of the same ages fell by 38 percent (96). Pipe and cigar smokers have been shown in the prospective stud- ies to have lung cancer mortality rates higher than those of non- smokers, although these are generally substantially lower than those of cigarette smokers (table 1). Retrospective Studies More than 30 retrospective (case-control) studies have been re- ported concerning the relationship of smoking and lung cancer. These studies are outlined in tables A3 and A4. Table A4 presents the percent of nonsmokers and of heavy smokers among both cases and controls as well as the relative risk ratios for all smokers. 240 Lez TABLE 1.—Lung cuncer mortality ratios (Actual number of deaths shown in parentheses)? SM = Smokers. NS = Nonsmokers. Prospective studies Author, Number year, and type collection Follow- Number Regular cigarette Pipe country, of Data up 0 smoking only cigar Inhalation Exsmokers Comments reference population years deaths (cigarettes/day) Hammond 187,783 Question- Bla 448 Pipe No data Bronchogenic 341/448 and white naire and SM . 443 NS ..... 1,00 (15) NS . 1.00 (15) (acluding adenocarcinoma) deaths with Horn, males interview. NS. 15 <10 . 8.00 (24) SM . 2.57 (18) Never smoked .......-- 1.00 microscopic 1958, in 9 ...10.50 (84) Cigar Previously <1 pack/day proof. In- U\S.A. States .. 23.40(117) NS ... 1.00 (15) Continuing weve ee 16.94 cludes those (120). ages ... 410.73 (397) SM ... 1.00 (7) Duration >10 years .. 1.51 smokers who Previously >1 pack/day also smoked Continuing .....-.-++: 46.21 pipes and Duration] 10 years ..17.79 without microscopic proof. Dolland Approxi- Question- 10 212 NS ....-- 1.00 (3) Pipe and Cigar No data Cigarette amokers Hill, mately naire and SM . 209 1-14 . 814 (22) NS.... 1.00 (3) NS coc cc cece cee ee cease 100 (3) 1964, 41,000 followup NS. 8 15-24 ...19.86 (53) Grams/day Continuing .....---+-- 18.29 (124) Great male of death >>25 | 82.48 (57) 1-14... 6.00 (12) Duration 20 years . 2.71 (2) Best, Approxi- Question- 6 3381 NS ..... 1.00 (7) Pipe No data { Refers 1966, mately naire and TSM . 324 <10 ....10.00 (67) NS ....1.00 (7) NS ..cccceee ce eeeee es 1,00 (7) to cure 1966, 78,000 followup NS. 7 10-20 ...16.41(204) 5M ... .4.35 (18) Ex-smokers 0: rent Canada male of death 20.0 ....17,81 (63) Cigar cigarettes only ...... 6.06 (18) cigarette (21). Canadian certificate. All ..... 14.20(245) NS ....1.00 (7) smokers veterans. SM ....2.94 (2) only. cre TABLE 1.—Lung cancer mortality ratios (cont.) (Actual number of deaths shown in Parentheses )1 SM = Smokers. NS = Nonsmokers. Prospective studies Author, Number year, and type Data Follow- Number Regular cigarette Pipe country, of collection up 0: smoking only cigar Inhalation Exsmokers Comments reference population years deaths (cigarettes/day ) Kahn U.S. male — Question- 8% 1,256 Pipe (Dorn), veterans naire and SM .1,178 NS ..... 1.00 (78) NS ....1.00 (78) No data 1966, 2,265,674 followup NS. 78 1-9 .... 5.49 (45) SM ....1.84 (17) NS ............ 1.00 (78) U.S.A, person of death 10-20 ... 9.91(303) Cigar Number of cigarettes/day: (189), years. certificate. 21-39 ...17.41(315) NS ....1.00 (78) 1-9 «ee. 0.95 (4) >39 ....23.93 (82) SM ....1.59 (6) 10-20 ......... 8.48 (39) All .....12.14(749) Pipe and cigar 21-39 ........, 9.83 (57) NS ....1.00 (78) >39 wo... 8.24 (19) SM ....1.66 (20) Hammond, 440,558 Interviews 4 Males Current cigarettes Pipe Males ICD code 1966, males by ACS 1,159 only NS ....1.00 (49) NS ........ 1.00 (49) 162 only, U.S.A. 562,671 volunteers. SM .1,116 Males SM ....2.24 (21) Slight ...... 8.42 (120) (118), females NS. 49 NS ..... 1.00 (49) Cigar Moderate ...11.45(311) 35-84 Females 1-9 .... 4.60 (26) NS ....1.00 (49) Deep ....... 14.31(141) years of 183 10-19 ... 7.48 (82) SM ....1.85 (22) Females age in 25 SM. 81 20-39 ... 18.14(881) Pipe and cigar NS ........ 1.00 (102) States, NS . 102 >40 ....16.61 (82) NS ....1,00 (49) Slight ....,. 1.78 (25) All ..... 9.20(719) SM ....0.90 (11) Moderate Females Deep i + 3.70 (45) NS ..... 1.00(102) 1-19 .... 1.06 (20) >20 .... 4.76 (50) AN ..... 2.20 (81) TABLE 1.—Lung cancer mortality ratios (cont.) (Actual number of deaths shown in parentheses)! SM = Smokers. NS = Nonsmokers. Prospective studies ere Author, Number Follow- year, and type Data up Number Regular cigarette country, of collection years of smoking only Exsmokers Comments reference population deaths (cigarettes/day ) Buell 69,868 Question- 3 304 NS .... 1.00 etal. American naire and <20 .... 2.80 1967, Legion- followup 20 ... 3.50 U,S.A. naires of death >20 .... 4.90 (49). 35-75 certificate. years of age and older. Hirayama, 265,118 Trained 1% 48 NS ..... 1.00 (3) Preliminary 1967, male and PHS SM. 40 1-24 .. 2.69 (29) report. Japan female nurse >25 . 5.68 (5) (125). adults interview 40 years and fol- ofageand lowup of older. death certificate. Weir and 68,153 Question- 5-8 368 NS -» 1.00 NS include Dunn, males in naire and +10 .... 3,72 pipe and 1970, various followup #20 ». 9.05 cigar U.S.A. occupa- of death >30 . 9.56 smokers (806). tions in certificate. All ..... 7.61 SM include California. ex-smokers. 1 Unless otherwise specified, disparities between the total number of deaths and the sum of the individual smoking categories are due to the exclusion of either occasional, miscellaneous, mixed, or examokers, TABLE 2.—Lung cancer mortality ratios for males by duration of cigarette smoking (Actual number of deaths are shown in parentheses) Age began cigarette smoking 35-54 55-69 70-84 35-84 25 or older .........--, 2.77 (5) 3.39 (12) 3.38 (3) 3.21 (20) 20-24 cece eee eee 5.83 (31) 11.11 (72) 12.11 (7) 9.72(110) TB-19 eee eee 8.71 (112) 13.06 (176) 19.37 (27) 12.81(315) IB eee cee ee eee 12.80 (35) 15.81 (57) 16.76 (9) 15.10(101) Source: Hammond,E. C. (118). These smoker-nonsmoker risk ratios range from 1.2 to 36.0 for males and from 0.2 to 5.3 for females. Although not presented in tabular form, the data concerning lung cancer and pipe or cigar smoking are similar to those found by the prospective studies mentioned above. However, a study by Abelin and Gsell (1) conducted on a rural Swiss population noted that an increased risk of lung cancer was present among heavy cigar and pipe smokers (as well as cigarette smokers) to a greater degree than previously reported. The authors suggest that their findings might be due to differences in either the amount smoked or the car- cinogenicity of Swiss and German cigars. The difference might also be explained by the greater use and more frequent inhalation of small cigars in Switzerland as compared to other countries where large cigars are more commonly smoked but rarely inhaled. Kreyberg (154), in a review of 887 cases of lung cancer in Norway, noted that pipe smokers showed an increased risk of lung cancer, although this risk was substantially lower than that for cigarette smokers. LUNG CANCER TRENDS IN OTHER COUNTRIES Several studies of particular interest are those in which the changing mortality from lung cancer has been investigated in countries in which cigarette smoking has become popular and wide- spread only in recent years. In those countries where accurate statistics for lung cancer mortality are available for both the pre- smoking and post-smoking periods, long-term trends can be studied in some detail. Two such studies have dealt with lung cancer mortality trends in Iceland. Dungal (83) noted in 1950 that lung cancer was a rare disease in Iceland and felt that this rarity could be explained by the relatively late onset of heavy tobacco smoking in the Icelandic population when compared to that of Great Britain and Finland. He observed that the annual per capita consumption of tobacco did not reach one pound in Iceland until 1945, while Great Britain and Finland passed that amount before 1920. In 1967, Thorarinsson, et al. (276) noted a sharp rise in the incidence of lung cancer in Ice- 244 60— SS Finiand Cy 50— 2 a BE Norway 40— 30— 8 20— 2 8 - to o é oe) 2 _/ LJ = 10 £ - © g e+ * B z ¢ 6— < 4— 3— 2— 1934-36 1939-41 1944-46 1949-51 1954-56 1959-61 1963-64 Calendar Years Ficure 1.—Lung cancer, Finland and Norway. Source: Kreyberg, L. (154). land after 1950 and found a correlation between that increase and the increasing sale of cigarettes in that country. Kreyberg (154) analyzed the lung cancer death rates of both Norway and Finland in relation to the use of tobacco in those two Countries over the past 100 years. Figure 1 shows the substantial difference in lung cancer mortality between the two countries. Kreyberg observed that cigarettes came into use in Norway in 1886 while the Finnish population (more closely allied to Russia socio- economically) was consuming more than 100 million cigarettes per year during the decade of the 1880's, Cigarettes remained scarce in Norway until after World War I, and this 30-year lag in consump- 245 TABLE 5.—Annual means of total lung cancer mortality and sex ratios for selected periods in Finland and Norway Finland Norway Year ——______ Males Females Males Females 1936-38 6... eee 192 33 34 30 Sex ratio ................ §.8:1 11:1 1968-65 6 eee 1,319 121 355 79 Sex ratio ................, 10.9:1 45:1 Source: Kreyberg, L. (154). tion behind that of Finland is reflected in a similar lag in total lung cancer mortality and sex ratios (table 5). HISTOLOGY oF LUNG TUMoRS A number of investigators have focused their interest upon the relationship of cigarette smoking to the varied histology of lung tumors. The major histological types of lung cancer include squa- mous cell (epidermoid) carcinoma, small and large cell anaplastic carcinomas, adenocarcinoma (including bronchiolar and alveolar types), and undifferentiated carcinoma (153). A review of these studies (table 6) indicates a closer relationship between cigarette smoking and epidermoid carcinoma than between cigarette smok- ing and adenocarcinoma (42, 113). The work of Kreyberg (153) in Norway, over the past 20 years, provides evidence of a specific histologic relationship. This inves- tigator noted that a clearer association is obtained if the various types of pulmonary carcinomas are grouped. Table A7 presents his groupings of the specific histologic types. Using this classification as a basis for analysis of lung cancer sex-ratios in N orway, Kreyberg has observed that Group I carcinomas are significantly more frequent among males while Group II carcinomas show an approximately equal] distribution among males and females. The author considers the recent rise in lung cancer in Norway to be a reflection of the increased prevalence of Group I carcinomas. Table 8 presents a summary of Kreyberg’s investigation concerning 793 male and female cases of lung cancer. Among both males and fe- males, the risk ratio among smokers is substantially higher for Group I types than for those of Group II. However, adenocarcinoma among males shows a risk ratio of 2.9, signifying a relationship with smoking. Kreyberg attributes the lower rates noted among females to their significantly lower consumption of tobacco in all forms. 246 Lez TaBLE 6.—Epidemiologic and pathologic investigations concerning smoking and the histology of lung cancer" (Actual number of cases shown in parentheses) Author, Number of year, persons and country, case selection Results Comments reference method Wynder 644 autopsies on Percent cases by histologic type and smoking history The percentage of chain and males with All lung cancers other than smokers in the general Graham, confirmed adenocarcinoma (605) Adenocarcinoma (89) population (7.6) was 1950, lung cancer. Nonsmokers 6... ee eee eee 1.3 10.3 significantly less than U.S.A. Light cigarette smokers ............0065 2.3 U4 among the patients with (816). Moderate 0... eee ee eee eee 10.1 15.4 adenocarcinoma. The Heavy oo e eee tee eter e eee 35.2 38.5 authors refrained from Excessive (occ ce cece eee eens 30.9 10.3 making any definite Chain 6 cece eee 20.3 18.7 conclusions due to the insufficient number of cases. Doll 916 male and 79 Percent patients with lung cancer by average amount smoked daily over 10 years No statistically and female cases Males significant difference Hill, with histologically Oat-cell or was found between 1952, confirmed Epidermoid (475) anaplastic (80%) Adenocarcinoma ($3) the amounts smoked by England lung cancer. Nonsmokers........ 0.2 (1) 0.7 (2) 6.1 (2) the patients in the (78). Smokers: different histological <5 cigarettes/day .. 2.9 (14) 3.9 (12) 6.1 (2) groups. Number of 35.6 (169) 36.3(110) 21.2 (7) proven adenocarcinomas W-25 cee eae 36.8 (175) 34.7(105) 48.5 (16) too small for ZB eee ee 24.4(116) 24.4 (74) 18.2 (6) conclusions. Females Oat-cell or Epidermoid (18) anaplastic (38) Adenocarcinoma (10) Males—105 unclassified Nonsmokers ......... 61.1 (11) $1.6(12) 50.0 (5) tumors. Smokers: Females-—13 unclassified <5 cigarettes/day .. 5.6 (1) 15.8 (6) 20.0 (2) tumors. 5-14 2 eee eee 22.2 (4) 23.7 (9) 10.0 (1) 15-25 wee eee ee eee 5.6 (1) 18.4 (7) see DOB eee ce eee eee 5.6 (1) 10.5 (4) 20.0 (2) 8re TABLE 6, E'pidemiologic and pathologic investigations concerning smoking and the histology of lung cancer’ (cont.) (Actual number of cases shown in parentheses) Author, Number of year, country, case seieelion Results Comments reference method Breslow 493 male and 25 Percent of patients with specific lung cancera by tobacco usage during the 20 years prior to study Nonsmokers include pipe et al., female cases and cigar smokers only. 1954, with histologically All lung cancers other than The authors conclude U.S.A. proven lung adenocarcinoma Adenocarcinoma Controls that cigarette smoking (42). cancer. (472) (518) appears to affect the 518 age and Nonsmokers oo... cece eee eas 5.9 24.4 development of sex-matched Cigarette smokers .............. 000000, 94.1 75.6 epithelial carcinoma controls. more than that of adenocarcinoma. Schwartz 430 male and Percent of smokers by histologic type and smoking history et al., female cases 1957, with histologically Epidermoid Anaplastic Unknown type Cylindrical + Difference France confirmed lung Cases .............. 96.0 97.0 100.0 significant (247). cancer. 4 matched Controls oo... 6... 29.0T 83.0T 96.0 at p<<0.05 level. control froups. Haenszel 158 female Relative risk for specified tumors (smokers/nonsmokers) 134 cases with final et.al., cases of histological 1958, lung cancer. Group I (Kreyberg) Adenocarcinoma determination. U.S.A. Adjusted for age and occupation. ............. 3.0t 1.19 + Difference from (118). unity significant at p=0.01. Haenszel 2,191 male Standardized mortality ratios Cases obtained from a and cases of 10 percent sample of Shimkin, lung cancer Epidermoid and undifferentiated lung cancer deaths in 1962, with adequate carcinomas Adenocarcinoma yg a. during 1958, U.S.A. histologic data. White males total ........ 00. ccc ccc cee ene 100 100 The authors noted an (112). Never smoked 6 18 absence of important Ex-smokers 34 46 differentials by <1 pack/day 123 116 histologic type. >1 pack/day ... 499 467 TABLE 6. pide miologic and pathologic investigations concerning smoking and the histology of lung cancer’ (cont.) (Actual number of cases shown in parentheses) Author, year, country, reference Cohen and Hossain, 1966, U.S.A. (58). Ashley and Davies, 1967, England (6). Ormos et al. 1969, Hungary (204). Number of persons and case selection method 417 male and female cases of lung cancer with histologic diagnosis 1439-63 at one hospital. Results Comments Percent cases by histologic type wid smoking history (number of smokers} 442 male and female cases of histologically diagnosed lung cancer, 113 male and female cases of histologically proven Jung vaAncer With adequate smoking information. Squamous Undifferentiated Adenocarcinoma Alveolar Nonsmokers ......... 1.0 (3) 10.0 (17) 23.0 (8) 20.0(1) Smokers... ..0...005 89.0 (183) 90.0(145) 60.0(20) Pereent eases by histologic type and smoking history Undifferentiated Squamous Adenocarcinom Nonsmokers ...........-- 2.8 (4) 2.5 (6) 3.4 (2) Pipe cece eee 9.9 (14) 9.9 (24) 1.7 (i) Cigarette 00.0... eee eee 87.3 (124) 87.6(211) 94.9(56) S10 day 2... ee eee ee 14.1 (20) 22.4 (54) 22.0(13) 10-20 .. 33.8 (48) 41.5(109) 33.9 (20) 21 30 12.0 (17) 21.6 (52) 16.9(10) 31-40 14.1 (20) 12.9 (31) 6.5 (5) 0 7.1 (10) 6.2 (15) 5.1 (3) The authors also noted that: 1. Adenocarcinomas were 215.8 times more common in women Only 1 percent of Kreyberg Group I eases were nonsmokers. ws The authors noted that cigarette smoking appears to be as strongly related to adcnocareinoma as to the other 2 lypes. Ashley’s data on tetal number of cigarette smokers are inconsistent with his breakdown of smokers into groups based on number of cigarettes smoked per day. Percent cases by histologic type and smoking history Croup land large ecll carcinomas 36.0 (9) 64.0(16) Group I 21.0(18) 79.0(68) Nonsmokers Smokers The author noted that the small number of eases allows for no definite conclusions. 6r7 Data obtained from patient. interview and other sources. Osz TABLE 8.—Tumor prevalence among males and females 35-69 years of age, by type of tumor and smoking category (Smokers constituted 85 percent of populations studied) Smoking category Expected Risk number ratio Sex and type of tumor Smoking Non- among among Total all methods smokers smokers } smokers Males Epidermoid carcinoma ........0. 0... cece eee e ence c ee aeeaens 434 431 3 17.0 25.4 Small cell anaplastic carcinoma ............. 00.0 ceec ence ccucceeus 117 116 1 5.7 20.4 Adenocarcinoma 0.1.0... ieee cee ene ee tence ence eenenenes 88 83 5 28.3 29 Bronchiolol-alveolar carcinoma ...... doe snes see an sees Carcinoid .................005 46 39 q 39.7 1.0 Bronchial gland tumor 0 685 669 16 90.7 TA Females Epidermoid carcinoma ............. 0.0 cc ccc ccc ccc cucceesetueeuge 12 9 3 15 12.0 Small cell anaplastic carcinoma ............. 0.0 cceceeees 8 5 3 15 6.6 Adenocarcinoma . 16... ccc cence teen neces 56 14 42 10.5 1.3 Bronchiololealveolar carcinoma 2.0.0.0... 0.0000 c ce cece cece cee ues — tae see an eee Carcinoid «1.0... cece cee cee teen cere crv breetnvenaenas 32 q 25 6.3 Ll Bronchial gland tumor ........... 0.0 .c ccc cccccaccecescescucee . 6 108 35 73 18.3 1.9 1+ Number that would be expected if incidence rate among smokers were Source: Kreyberg, L. (154) equal to that of nonsmokers.