Tobacco Documents Online

Acta Otolaryngol (Stockh) 1997; Suppl 527:160-164 Epidemiology of Laryngeal Cancer: Results of the Heidelberg Case-Control Study HEINZ MAIER and MAT~HIAS TISCH From the Department of Otorhinolaryngology/Head and "Neck Surgery, Armed Forces Hospital Ulm, Germany Maler H, Tlseh M, Epidemiology of laryngeal cancer. Results of the Heidelberg case-control study. Aeta Otolaryngol (Stoekh) 1997; Suppl 527: 160-164. "' Squamous cell carcinoma of the larynx is a multifactorial disease which predominantly is found in males aged 50-70 year~. Chronic consumption of alcohol and tobacco independently increase the relative risk of this type of cancer in a dose-dependent manner. Low educational standards and occupational training are associated with high risk. The majority of the cancer patients are blue collar workers who are exposed to a variety of hazardous working materials like polyeylie aromatic hydrocarbons, cement dust, metal dusts, asbestos, varnish, lacquer, etc. Environmental exposure to airborne carcinogens like fossil fuel single stove emissions may increase the relative risk of laryngeal cancer. The regular consumption of fruit, salad and dairy products which contain considerable amounts of tumor protective mieronutrients may decrease the risk of laryngeal cancer. Key words: epidemiology, laryngeal cancer, risk factors. INTRODUCTION Squamous cell carcinoma of the larynx is the most common malignant tumor in the upper aerodigesfive tract. The reported incidence rates in European coun- tries range from 17.2 in Navarra, Spain to 2.8 in Sweden (1). Incidence among males in most popula- tions is more than 10 times higher than among fe- males (1). Laryngeal cancer is much more common among blacks than among whites (1). Diagnosis of laryngeal cancer in patients younger than 40 years is rare. In Germany the disease manifests most often within the 50-65 years age period (2). Laryngeal cancer is a multifactorial disease, which is firmly linked to life-style and environmental fac- tors. Recently, a considerable amount of epidemio- logic evidence has been accumulated to implicate chronic consumption of alcohol and tobacco, envi- ronmental factors, occupational exposure to carcino- genic or cocarcinogenic factors, diet and social status in the etiology of this type of cancer. This paper is a case-control study of risk factors conducted in a group of German patients with laryn- gral cancer which will be supplemented by a short review of the 1.iterature. MATERIAL AND METHODS The study was pertbrmed at the Department of Otorhinolaryngology/Head and Neck Surgery of the University of Heidelberg. It comprised 164 male pa- tients with histologically proven squamous cell car- cinoma of the larynx. We attempted to recruit all patients who attended the hospital for treatment and follow up examination within the time periods be- tween February 1, 1988 and May 30, 1988, and between November 1, 1988 and May 1, 1989. To © 1997 Scandinavian University Press. ISSN 0365-5237 reduce the possibility of bias and/or selection, only patients for whom the interval between diagnosis of the tumor and interview was not longer than 3 years were eligible. A total of 656 male subjects without evidence of cancer served as controls and were se- lected randomly from the outpatient dirties of the Departments of Otorhinolaryngology (n = 328) and Internal Medicine (n = 328). Cases and controls were matched for age an~d residential area using a 1:4 matching design. All interviews were conducted by the same person using a structured questionnaire. The interviews took approximately 30-60 rain and covered alcohol consumption, smoking habits, di- etary habits, social status, occupational history and previous medical conditions. Tobacco consumption was described by the term "tobacco year" (TY), which was defined as a daily consumption of 20 cigarettes or 4 cigars or 5 pipes for I year. Alcohol consumption was described as total intake of ethanol (g/day). The calculation of this value was based on an average ethanol content of 35 g/1 in beer, 80 gfl in wine, and 320 g/1 in liquor. The analysis of the data was performed by means of the statistical.package SAS. Risk estimates and corresponding confidence intervals were calculated by means of logistic models of regression using the pro- cedures LOGIST and MCSTRAT. Results are pre- RESULTS AND DISCUSSION Among the 164 cases there were 33% plain glottie cancers and 32.9% plain supragiottie cancers. In 34.1% of the cases a clear differentiation between supragiottic and glottie cancer was not possible. The mean age of the cases was 58.15 years (minimum age: 39.9 years, maximum age: 85.8 years). The average ~ ardcle i~ for indi~du~l u.~ ~b/~nd ma~ not b~ further reproduced or stored eleetr~e~/ly without ~itle~a permission from thn a~l~,right holder. Unauthorized r~preduct~c~n may r~dt in financial z~nd ol!ner pea'mlifie~. (e) SCANDINAVIAN UNIVERSITY PRESS

Epidemiology of laryngeal cancer 161 1 i i i 1 1 1 i ! l 1 1 1 Table I. Distribution of never smoker, previous smoker and active smoker (%) Cases Controls Never smoker 4.3 30.0 Previous smoker 77.4 41.0 Active smoker 18.3 29.0 p < 0.00 I. age in patients with glottic cancer at the time of diagnosis was higher (57.1 years) compared to those with supraglottic cancer (49.2 years). Tobacco consumption Of the cases, 4.3% were never smokers, 77.4%0 were previous smokers and 18.3% were still active smok- ers. Among the controls 30% had never smoked, 41% were previous smokers and 29% were active smokers (Table I). The smokers among the study cases in 96.5% were cigarette smokers (controls 91.5%), 30% of them preferring cigarettes without filtertip (con- trols 12.2%). The mean age at which smoking was begun by the cancer patients was 18 years (controls I9.1 years). The mean tobacco consumption among the cases was 47.9_+33.1 TY (controls 23.3_+27.3 TY). The tobacco-associated relative risk (RR) is shown in Table II. Compared with the referent cate- gory., which included tobacco consumers with less than 5 TY, there was a clear dose-response relation- ship that displayed a 9.5-fold increase in risk (ad- justed for alcohol consumption) in heavy smokers (more than 50 TY). There is general agreement that cigarette smoking is the major risk factor for laryngeal cancer. The tobacco-associated cancer risk has been explained by the fact that tobacco smoke contains more than 30 different carcinogenic compounds, for example poly- cyclic aromatic hydrocarbons, and tobacco-specific nitrosamines some being capable of local carcino- genic action (3)2 Like numerous former studies in other countries (4) the present study also clearly demonstrates a dose-dependent increase of laryngeal cancer risk in smokers. A comparatively high per- Table III. Frequency of alcohol consumption and pre- ferred alcoholic beverages (%; repeated terms possible) Cases Control Daily 85. I 64.4 Sometimes/month 3.1 13.1 Seldom 7.5 21.9 Never (p < 0.001) 4.4 0.6 Beer (p < 0.001) 92.7 79.6 Wine (p < 0.11) 35.7 42.9 Liquor (p < 0.001) 32.5 12.9 centage of cancer patients smoked nontilter cigarettes. Consumption of nonfilter cigarettes means a higher exposure to tumorigenic components and indeed causes an additional increase of risk (5). Alcohol A daily consumption of ethanol was reported by 85.1% of the cancer patients and by 64.4% of the controls (Table III). The preferred beverages in the cancer patients were beer and liquor (Table III). The mean daily alcohol consumption among the cancer patients was 65.1 ±49.9 g (controls 29.1 ±31.1 g). The risk ratios associated with alcohol consumption are shown in Table IV. The referent category in- cluded subjects with a daily consumption of less than 25 g/day. There was a dose-response relationship displaying a 14.7-fold increase in risk (adjusted for tobacco consumption) for heavy drinkers (more than 100 g/day). Clinically a link between chronic alcohol consump- tion and the risk for laryngeal cancer has been ob- served for decades and this observation has been supported by a number of epidemiologic studies (4). Initially alcohol has been described as a risk enhancer in smokers. More recent studies, however, provided evidence that alcohol acts as an independent risk factor for laryngeal cancer (6, 7). Also in the present study we observed a dose-dependent increase in risk, which remained enhanced even after adjustment for tobacco consumption. It seems likely that the cancer risk is linked mainly to the total intake of alcohol and Table II. Tobacco associated risk of laryngeal cancer Tobacco Relative risk Relative risk consumption (crude) 95%, C.I. (adj. for alcohol) 95%, C.I. rO < 5 1.0 - 1.0 - 03 5-19 3.6 [1.7-8.0] 4.0 [1.7-9.2] O~ 50-74 6.6 [3.3-13.4] 6.3 [3.0-13.3 ~ 75-99 9.6 [4.7-19.7] 7.8 [3.6-16.7] -,O > 100 12.7 [6.4-25.1] 9.5 [4.6-19.6] O~

162 H. Maier and M. Tisch Table IV. Alcohol associated risk of laryngeal cancer Alcohol consumption Relative risk Relative risk (g/day) (crude) 95%, C.I. (adj. for tobacco) 95%, C.I. < 25 1.0 - 1.0 - 25-49 2.4 [1.5-4.0] 2.3 [1.3-3.8] 50-74 4.1 [2.4-6.9] 3.4 [1.9-6.1] 75-99 7.3 [3.7-14.4] 6.6 [3.2-13.6] > 100 18.7 [19.7-35.8.]. 14.7 [7.4-29.4] not to the type of preferred beverage consumed. Due to methodological reasons we did not study the inter- action of alcohol and tobacco on the risk for laryn- geal cancer. However, previous investigations on risk factors of head and neck cancer provided evidence that alcohol interacts in a synergistic way with to- bacco smoke (4). The mechanisms which underlie alcohol-related cancers are not completely elucidated. At present we know that alcohol is not a carcinogen, but acts as a co-carcinogen that is capable of effecting carcinogen- esis locally and systemically via different mechanisms and at different stages during initiation and promo- tion (8). Further chronic alcohol consumption is of- ten associated with malnutrition and a depletion of tumor protective vitamins and minerals (8). Diet Patients with laryngeal cancer reported a lower con- sumption of fruit, salad and dairy products (Table V) whereas the consumption of eggs and innards (kid- ney, heart, liver) was higher. Calculation of risk estimates revealed that regular consumption (more than once weekly) of dairy products, fruit and salad was associated with a protective effect concerning laryngeal cancer (Table VI). There is sufficient evidence that deficiencies in vita- mins A, C, E, beta-carotine, riboflavin and iron, zinc, and selenium increase the risk of larynge~,l cancer (9). These micronutrien~s have been shown to act as antioxidants and/or differentiation inducers and thus are supposed to inhibit carcinogenesis at different stages. The present study revealed a low intake of Table V. Consumption of different food items (more than once weekly) by patients with laryngeal cancer and control subjects Food Cases Controls p-Value Milk and milk products 51.8 63.2 0.007 Eggs 36.6 24.2 0.001 Fruit 71.9 81.3 0.008 Salad 82.3 87.2 0.105 fruit, salad, and dairy products in laryngeal cancer patients. These food items contain considerable amounts of the above mentioned micronutrients. The higher consumption of eggs and innards in the cancer group fits with a pattern of laryngeal cancer patients being relatively unconcerned about health practices in general, which again might be at least in part due to lower education and social status. Social status School education and occupational training in the patients with laryngeal cancer and the control group are shown in Table VII. Only 10.1% of the cancer patients had completed high school (controls 25.2% ). Technical college, college or university had been com- pleted in 8.2% of cases but by 31.5% of controls. 34.8% of the laryngeal cancer patients in comparison to 20.3% of control subjects had lived for more than 40 years in small apartments. The risk estimates calculated from these data revealed a significantly increased laryngeal cancer risk for subjects with a low educational level (RR = 2.4; C.I. 1.3-4.5; adj. for alcohol and tobacco consumption) and with a low occupational training (R.K= 4.2; C.I. 2.2-8.2; adj. for alcohol and tobacco consumption). Previous epidemiologie studies have reported asso- ciations between socio-economie status and laryngeal cancer (10, 11). Our study supports these observa- tions, displaying a lower educational level and a lower occupational training in laryngeal cancer pa- tients in comparison to randomly selected control subjects. The poor housing conditions observed in the cases is another indicator for their affiliation with the lower social classes. We believe that this observation above all is due to poor health care and especially to smoking, drinking and dietary habits in these social classes. However, as well as this a more intensive exposition to environ- mental and occupational carcinogenic factors has to be discussed. Occupation The majority of the laryngeal cancer patients were blue collar workers exposed to a variety of hazardous

Epidemiology of laryngeal cancer 163 Table VI. Relative risk of laryngeal cancer associated with the regular consumption (more than once weekly) of different food items (adj. for alcohol and tobacco consumption) p -Value 0.18 0.08 0.08 Food RR 95%, C.I. p-Value RR 95%, C.I. -- Milk and milk-products 0.6 [0.4-0.9] 0.008 0.8 [0.6-1.1] Fruit 0.6 [0.4-0.9] 0.010 0.7 [0.4-1.1] Salad 0.7 [0.4- I. 1] 0.110 0.6 [0.4-1.1] ~¢ork quen hydr~ and 1 cons1 expo for consl trois were • espe~ RR t (C.I. tion) (C.I. tion). Table VIIa. School education and occupational training of patients with laryngeal cancer and control subjects Cases Controls (n -- 158) (n = 648) No training 13.3 9.1 Apprenticeship 65.8 49.4 Training 12.7 I0.0 Technical college 5.7 18.1 RR (C.I.) College 1.9 6.8 (crude) University - 1.5 Cement dust • p < 0.001. PAH Table VIIb. Relative risk of laryngeal cancer with Wood dust, general Pine wood dust education level and occupational training (adj. for alco- Paint, lacquer hol and tobacco consumption) Dust general RR p-Value 95%, C.I. Educational level (no 3.14" 0.02 [1.13-4.06] final examination primary/secondary Fossil fuel single school) stove for a period Occupational training 3.8'1" 0.001 [1.94-7.45] of (no training, 0-20 years 1.0 apprenticeship) 20-40 years 1.2 >40 years 2.5 • RR = 1.0; O-levels, technical schools, A-levels. > 40 years 2.0 i" RR = 1.0; technical college, college, university. working material (Table VIII). Cases were more fre- quently exposed to cement dust, polycyclic aromatic hydrocarbons (PAH), pine wood dust, and varnish and laquer. After adjustment for alcohol and tobacco consumption the risk estimates calculated from the exposure data ranged from 1.8 for cement dust to 2.7 for polycylic aromatic hydrocarbons (Table IX). Among the cancer patients 23.2% had worked as construction workers for more than 10 years (con- trois 9%). Furthermore, the laryngeal cancer patients were more frequently employed in the textile sector, especially in the manufacture of lehther products. The RP, for employment as a construction worker was 2.3 (C.I. 1.3-4.0; adj. for alcohol and tobacco consump- tion) and for employment as a textile worker was 3.5 (C.I. 1.3-9.4; adj. for alcohol and tobacco consump- Within the last few decades a number of epidemio- logical studies have provided substantial evidence for an association between a variety of occupational agents and head and neck cancer, especially laryngeal cancer (12). Similar to the present study they indicated an increased risk for the blue collar workers, most of them being exposed to various inhaled inorganic/or- ganic chemical agents. Besides the working materials mentioned in our study, exposure to asbestos, ionizing radiation, various metal dusts, diesel exhausts, sul- phuric acid mists, and mustard gas have also particu- larly been associated with laryngeal cancer. One might argue that the increased risk observed among blue collar workers may be explained mainly by the known high tobacco and alcohol consumption within this social group. However, this argument, at least in part, can be refuted since in most of the studies mentioned above the risk estimates remained high after adjustment for tobacco and alcohol. On the other hand it seems to be very likely that chronic tobacco and alcohol consumption combined with ex- posure to occupational carcinogens might increase Table VIII. Relative risk of laryngeal cancer associ- ated with exposure to working materials RR (C.I.) (adj. for tobacco and alcohol) 2.8 (1.7-4.6) 1.8 (1.0-3.2) 2.7 (1.4-5.5) 2.7 (1.2-6.1) 2.0 (1.1-3.3) - 2.1 (1.2-3.7) 2.0 (1.1-3.9) 2.3 (1.3-4.1) 2.3 (1.1-4.5) 5.9 (3.0-11.8) 6.2 (2.6-14.9) Table IX. ORs for the variables fossil fuel single stores RR 95%, C.I. p-Value [0.8-1.9] 0.5 [1.5-4.1] 0.0003 [1.1-3.5] 0.02 -!

164 H. Maier and M. Tisch Table X. Larynx collective: subjects general environmental situations Situations (years) Cases ~'A) Controls (%) p-Value Air pollution on the job (>20) 20.7 15.4 0.10 Traffic jams on the way to work (>20) I4.0 High traffic emissions in residential areas (>20) 23.5 Outdoor air pollution in residential areas (>20) 13.0 Household heating with fossil fuel stoves (>40) 34.8 Cooking with fossil fuel stoves (>40) 20.1 19.2 0.10 0.90 - 12.8 0.90 20.3 0.0001 14.6 0.08 the cancer risk in a more than additive manner, as it has been shown for smoking and asbestos exposure in lung cancer patients (13). Environment The general environmental situation of the laryngeal cancer patients and the controls in our study are shown in Table IX. Significantly high percentages regarding the use of fossil single stoves and cookers could be observed among the cases. 34.8% of the laryngeal cancer patients used fossil single stove heaters for more than 40 years (controls 20.3%). Burning lignite, hard coal, oil or wood in such single stoves causes an emission of comparatively high con- eentrations of carcinogenic combustion products like benzo[a]pyrene being partly adsorbed to particles into the indoor air (approx. 1000 times higher than with central heating). The RR of laryngeal cancer estimated from these data increased significantly with exposure time (Table X), and after more than 40 years reached a value of 2.0 (C.I. 1.1-3.5; adj. for alcohol and tobacco). Besides occupational exposures the environment might also be a source of airborne carcinogens. This seems to be especially true for indoor air pollution. According to the WHO the indoor air pollution by emission products of fossil fuel single stoves is a main factor of mortality in diseases of the respiratory tract in third world coun- tries (14). While until now indoor air pollution was only associated with lung cancer our study clearly indicates also that the risk of laryngeal cancer might be inc"eased. In future studies also the role of passive smoking, formaldehyde 'exposure and other factors of indoor air pollution as risk factors for head and neck cancer need to be studied. Besides indoor air pollu- tion, air pollution in general has been noted as a possible risk factor for laryngeal cancer. Accordingly in some countries a t.endeney for the disease to be more common in urban than in rural regions was reported (1). REFERENCES I. Muir C, Waterlaouse J, Mack T, Powell J, Whelan S, eds. Cancer incidence in five continents, Vol. 5. Lyon: International Agency for Research on Cancer, 1987. 2. Krebsregister Saarland, Morbidit.~t und Mortalit~t an b6sartigen Neubildungen ira Saarland 1991, Jahres- bericht Statistisches Landesamt Saarland (Hrsg.), Son- derheft 175, 1994. 3. Hoffmann D, Melkian A, Adams J-D, Brunnemann KD, Haley ND. New aspects of tobacco carcinogene- sis. Carcinogenesis 1985; 8: 239-56. 4. Maier H, Dietz A, Gewelke U, Heller WD, Weidauer H. Tobacco and alcohol and the risk of head and neck cancer. Clin Invest 1992; 70: 320-7. 5. Wynder EL, Stellman SD. Impact of long-term ~ter cigarette usage on lung and larynx cancer risk: A case-control study. J Natl Cancer Inst 1979; 62: 471-7. 6. Tuyns AJ, Esteve J, Raymond R, Berrino F, et al. Cancer of the larynxpaypopharynx, tobacco and alco- hol: IARC international ease-control study in turin and Varese (Italy), Zaragosa and Navarra (Spain), Geneva (Switzerland) and Calvados (France). Int J Cancer 1988; 41: 483-91. 7. Brugere J, Guenel P, Leclerc A, Rodriguez J. Differen- tial effects of tobacco and alcohol in cancer of the larynx, pharynx and mouth. Cancer 1986; 57: 391-5. 8. Seitz HK, Simanowski UA. Alcohol and caminogene- sis. Annu Rev Nutr 1988; 8: 99-119. 9. Graham S, Mettlin C, Marshall J, Priore R, Rzepka T, Shedd D. Dietary factors in the epidemiology of cancer of the larynx. Am J Epidemiol 1981; I13: 675-80. 10. Olsen J, Sabroe S. Occupational causes for laryngeal cancer. J Epidemiol Community Health 1984; 38: 117- 121. 1 I. Elwood JM, Pearson JCG, Sldppen DH, Jackson SM. Alcohol, smoking, social and occupational factors in the etiology of cancer of the oral cavity, pharynx and larynx. Int J Cancer 1984; 34: 603-12. 12. Maier H, de Vries N, Snow GB. Occupatio. nal factors in the etiology of head and neck cancer. Clin Otolaryn- gol 1991; 16: 406-412. 13. Hammond IC, Selikoff I J, Seidman H. Asbestos expo- sure, cigarette smoking and death rates. Ann NY Acad Sci 1979; 330: 473-90. 14. World Health Statistics Annual 1983, Geneva: World Health Organisation, 1983. Address for correspondence: H. Maier Department of Otorhinolaryngology/Head and Neck Surgery Armed Forces Hospital ULna, Germany Oberer Eselsberg 40 D-89081 Ulm Germany