Cancer”; R_ates and Risks sxsgu pue $6123 .Iaoueg 3rd edition, 1985 Cancer Rates and Risks 3rd edition, 1985 Harriet S. Page. Office of Cancer Communicatwons Ardyce J Asure‘ Demographic AnalySIS Section National Cancer Institute ‘ s_ _ MAY 1 D 117‘“; U 3 DEPARTMENT OF Pubhc Hualth Sen/woe NIH Publication No. 85-691 HEALTH AND HUMAN Nanona‘ \nsmutes of Health Apnl 1985 SERVICES Acknowledgements Much of the data in this third edition of Cancer Rates and Risks came from Stu-willance, Epidemi- ology and End Results: Incidence and Mortality Data. [973—77, prepared by the Demographic Analysis Section of the Division of Cancer Preven- tion and Control. and published as NCI Mono- graph 57 in 198]. Some later SEER data are also included. Cancer Mortality in the United States: 1950—1977. prepared by the Epidemiology and Bio- statistics Program of the Division of Cancer Etiology and published in 1982 as NCI Monograph 59. was another major source of data. Other sources are cited in the text and legends. Invaluable background assistance was provided by Gilbert W. Beebe. Aaron E. Blair. William A. Blattner. William J. Blot, John D. Boice. Louise A. Brinton. Kenneth P. Cantor. Joseph F. Fraumeni. Jr.. Mark H. Greene. Patricia S. Hartge. Robert N. Hoover, Arlene F. Kantor. Charles E. Land, Thomas J. Mason. Robert W. Miller. John J. Mulvihill. Linda Pottern, Joseph Scotto. Debra T. Silverman. Terry L. Thomas. Deborah M. Winn. and Regina G. Ziegler. ofthe Division of Cancer Etiology. and by David P. Byar. Roger R. Connelly. Peter Greenwald. David P. Levin. Earl S. Pollack. and John L. Young. Jr.. of the Division of Cancer Prevention and Control. Science-writing interns Leslie Fink. Michele Gauthier. and lna Silverman. of the Office of Can- cer Communications. wrote some ofthe sections. Amelia Champion provided editorial expertise. Table of Contents 1 Rates Page International Cancer death rates in 20 countries for: All sites 6 Breast 8 C olorectal 9 Esophagus ]() Larynx 1 1 Lung 12 Oral | 3 l’rostate | 4 Stomach 15 Uterus 16 Cancer incidence International range of cancer incidence 17 United States Cancer incidence Age-specific. all sites. by age and race 20 By age group. all races. both sexes 21 Incidence of 10 most common cancers by sex among whites. blacks. and Hispanics. (1973—77) 22 Changing patterns among whites and blacks. men and women (1969—1977) 24 Cancer death rates Changes for all cancers by age group. all races and both sexes 27 Changing patterns for eight major cancers. men (1973—77) 28 Changing patterns for eight major cancers. women (1973—77) 29 Five-year survival rates Changes for 19 sites among whites 30 Changes for 19 sites among blacks 31 Changes among white children underage 15 32 Costs of Cancer Costs of cancer. heart disease. stroke. and injuries. 1980 33 Medical care expenses in 1980 34 Earnings lost in 1977 35 2 Risks Risk factors Air and water 38 Alcohol 42 Diet 44 Drugs 48 Familial 53 Ionizing radiation 55 Occupation 58 ' Solar radiation 66 Tobacco 68 Viruses 70 Risks for major cancers Biliary tract 73 Brain 75 Breast 77 Childhood 79 Colon and rectum 83 Esophagus 87 Hodgkin's disease 89 Leukemia 9| Liver 94 Lung and larynx 96 Melanoma 99 Multiple myeloma 101 Non—Hodgkin‘s lymphoma 103 ()ral cavity and pharynx 105 ()vary 107 )anc reas 109 Prostate l 10 Skin (nonmelanoma) lll Stomach l 14 Testis l 16 Urinary tract 118 Uterine cervix 121 Uterine corpus 123 Glossary 126 References 128 RRRRR Cancer Death Rates in 20 Countries All Sites Death rates per 100,000 pop- ulation for all cancers in 20 countries, 1976—77, age-ad- justed to U.S. standard 1970 population. From Wor/d Health Statistics Annual 1979780 as adapted by Amer/can Cancer Society 7983 Country Rate 2698 2616 256.9 2515 2487 2448 2368 2321 2296 2291 2210 2136 2134 2112 2105 1977 1975 1880 1867 1705 Scotland Netherlands (Neth) Hungary England Austria Germany FR (W Ger) Swnzerland (Swrtz) Denmark Northern Ireland (N Ire) Hong Kong (H K J New Zealand (N Zea) Untted States (U 8) Poland Canada Australia Chile Sweden Norway Japan Israel Cancer occurs throughout the world: no country. no population is free of it. Cancer does not. how- ever. occur with the same frequency in all coun- tries: there are wide variations in total cancer death rates from country to country. There are also wide variations among different countries in the death rates for specific cancers. The bar graphs that follow show some ofthese variations. The cancer death rates that are displayed in these 10 graphs were derived from the number of cancer deaths per 100.000 population in each coun- try for the years l976—l977. The 20 countries were chosen because all have major populations with good. nationwide reporting systems. and each of the countries reports to the World Health Organi— zation. To make meaningful comparisons between countries. the rates were all age-adjusted to a stan- ard age distribution. Cancer deaths occur most often in older age groups. but not all countries have the same age proportions. Age—adjusting cor- rects for these differences. Male C. O 50 100 150 200 250 300 Country Denmam ScoHand Hungaw England Aumna VV Ger ChHe N.Zea N.He th. Israel Sweden U.S. Canada Swnz Norway Australia Poland HK. Japan Ram 1709 1658 1636 1560 1553 1548 1538 1507 1504 1426 1413 1409 1363 1353 1342 1312 1292 1264 1253 1087 Cancer death rates indicate only the number of persons who die from cancer in a given year; they do not necessarily coincide with the incidence of new cases of cancer in that year. But tumor regis- tries are needed to monitor cancer incidence. and not all countries have them. Most countries do have systems for reporting vital statistics. Cancer death rates thus provide a way to make interna- tional comparisons. (The U.S. cancer death rates in these charts are for all races.) The first graph shows death rates for all cancers. by sex, among the 20 countries. The nine graphs following show death rates for specific cancers. O 50 100 150 200 250 300 Cancer Death Rates in 20 Countries Breast Death rates per 100,000 pop- ulation for female breast can- cer in 20 countries, 1976—77, age-adjusted to U.S. standard 1970 population. From: World Health Statistics Annual 7979—80 as adapted by American Cancer Socrety 7983 Country Rate Denmark 33 8 Engbnd 336 N.he 330 Nah 315 Scotland 31 3 N. Zea 30 O Israel 29 2 SWIIZ. 28 6 Canada 284 LlS. 273 W Ger 251 Australia 24 8 Aumna 247 Hungaw 238 Sweden 235 Norway 22 6 Pomnd 163 ChHe 141 HK 107 Japan 57 Female breast cancer is rare in much of Asia. but it has been the most common cause of cancer death in North America and most of Europe. In the United States. lung cancer is overtaking breast cancer as the most common cause of cancer death among women. Breast cancer is very rare in men. The highest breast cancer death rates have been found among Caucasian women in Hawaii and in British Columbia. both with death rates greater than 80 per 100.000 population. Among the coun- tries listed below. the highest death rate is seen in Denmark, followed by Great Britain. The lowest are seen in Hong Kong and Japan. Low death rates also prevail in Africa. The breast cancer death rates for Chinese and Japanese women in the United States are three times higher than in Singa- pore or Japan. but not so high as the death rates among whites in the United States. The breast can- cer death rate among US. black women is lower than among white women. but is increasing. ""lIIIIIIIIIIIIIIF g o O N O 40 60 80 100 Colon and Rectum Risk factors, incidence, and death rates vary for cancers of the colon and the rectum. but the two are often considered together, as colorectal can- cers. They are considered diseases of economically developed countries: death rates are highest in western Europe, North America, and New Zea— land, and lowest in Asia, Africa, and most Latin American countries. In this country, deaths from colorectal cancers rank second only to lung can- cers in the total number of deaths. but death rates vary widely by geographic area in this country. Death rates are highest, for example, in Northeast urban areas and lowest in the South and South- Death rates per 100,000 pop- ulation for colorectal cancer west. in 20 countries, 1976—77, age- adlusted to U.S. standard 1970 population. From: World Health Statistics Annual 1979-80 as adapted by American Cancer Society 1983. Country Rate Male Country Rate N, lre, 33.5 N. Zea 300 Austria 33.4 Scotland 26.5 N. Zea. 330 Denmark 259 Scotland 327 N, lre 25.5 W. Ger 32.4 W. Ger. 24.8 Denmark 324 Canada 23.0 England 296 England 22.8 Canada 28.6 Australia 22.7 Hungary 28.5 Hungary 22.6 Australia 281 AUSilI’d 22.5 Neth 269 Neth. 21.1 SWItZ. 26.6 U.S. 20.2 US. 26.2 Norway 19.4 Sweden 25.0 Sweden 18.6 Norway 234 SWItz. 176 Israel 214 Israel 16.6 H.K. 16.9 H.K. 11.9 Japan 15.0 Japan 111 Poland 14.5 Poland 11.1 Chile 9.9 Chile 10.0 1 O 20 40 60 80 100 Cancer Death Rates in 20 Countries Esophagus Death rates per 100,000 pop- ulation for esophageal can- cer in 20 countries, 1976—77, Cancer of the esophagus is characterized by wider variations in incidence and death rates than any other cancer. There are geographical “pockets" of very high death rates in parts ot‘the Soviet Union. lran. China. and France. In some ofthese pockets the male-female ratios are very high. Among the countries listed below. male death rates are highest in Hong Kong and Chile. and lowest in Israel. Rel- atively high death rates from esophageal cancer are also seen among women in Hong Kong and Chile. In the United States. the male death rate from esophageal cancer is 3.6 times higher than for women. A pocket has been found in this country among black men in Washington. I).C.. who have a death rate of 28.6 per l00.()()(). age-adjusted to U.S. standard 1970 population. From World Health Stat/sires Annual 19797780 as adapted by Amer/can Cancer SOC/(fly 7983 Country H K Cnlle Japan Swnz Scotland England N Zea Australia U S N lre Poland Austria Canada W Ger Netn Denmark Sweden Hungary Norway Israel Rate 143 128 95 9.4 92 73 66 58 54 54 53 St 49 49 4,4 4.2 38 3.8 33 31 Male Country Chile Scotland H K England N Ire N Zea Australia Japan lsrael Neth Denmark U 8 Canada Sweden Swrtz Poland W Ger Norway Austria Hungary Rate 60 53 42 39 36 3.2 25 22 20 l6 16 t5 15 72 12 12 1.0 09 0.7 05 Ill l’\) O 40 60 80 700 60 80 700 Larynx Death rates per 100,000 pop- ulation for cancer of the larynx in 20 countries, 1976—77, age-adjusted to U.S. standard 1970 population. From' World Health Statrstrcs Annual 7979—80 as adapted by Amer/can Cancer Socrety 7983. Death rates from cancer of the larynx. or “voice— box." vary widely throughout the world. The high- est death rates in the world are found among men in Sao Paulo. Brazil (I4. I) and in parts of Spain (13.6). In the United States. the highest laryngeal cancer death rate is now found among black men in the San Francisco Bay area. Historically. laryngeal cancer has been a male disease. In the table below. for example. the highest male death rate. among Hungarian men. is ID times greater than the high- est female death rate. and the lowest male death rate. in Sweden. is 2 times greater. There are indi- cations. though. that death rates from laryngeal cancer are now rising in women. Country Rate Male Country Rate Female Hungary 7.4 I N, Ire. 0.7 I Poland St I Scotland 0 6 I H K 5 o I Israel 0 6 | Austria 4 6 I Chlle 0.5 I Swntr 3 9 I H K o 5 I lsrael 3 2 I Hungary 0 5 I W. Ger 3 2 I Poland 05 I Canada 3 t I U S O 4 I u s 2 9 | Canada 0 4 | Australla 2 9 I England 0 4 I ChIlO 2 9 I Denmark 0 3 I N Ire, 2 6 I Austria 0.3 I Netn 2 4 I Japan 0 3 I Denmark 2 3 I Swttz O 3 I England 2 3 I Neth 0.2 I N Zea 2 3 I Australia 0 2 I Scotland 2 3 I W Ger 0 2 I Japan l 8 I N Zea O 1 I Norway 1 4 I Norway 0 l I Sweden l 3 I Sweden 0 1 I O 20 40 60 80 100 0 20 4O 60 80 100 Cancer Death Rates in 20 Countries Lung Death rates per 100,000 pop- ulation for lung cancer In 20 countries, 1976—77, age-ad- justed lo u.s. standard 1970 population. From: World Health Statistics Annual 1979—80 as adapted by Lung cancer is a major cause of death in most Western countries, particularly, as this graph shows, among men. (Cancers ofthe lung, bron- chus, and trachea are grouped together in the World Health Organization reporting system.) In the United States, lung cancer is the leading cause of cancer death among men and is expected. in 1984 or 1985, to outstrip breast cancer as the lead- ing cause of cancer death among women (Horm and Asire. 1982). Lung cancer is one ofthe three most common cancers in men throughout the world, and probably outranks stomach cancer as the most common cancer in men (Fraumeni and American Cancer Society 1983. African countries. Country Rate Male Country Scotland 1085 ’ ' H.K. Neth. 97.7 Scotland England 965 England N. Ire. 71.7 U.S. US. 68.1 N, lret Austria 676 Denmark H.K. 65.6 N. Zea. N. Zea. 65.6 Canada Hungary 650 Hungary W. Ger. 64.0 Israel Canada 62.7 Australia Denmark 82.0 Austria Switz. 61.7 Japan Australia 61.6 Sweden Poland 60.2 Poland Sweden 33.4 Chile Israel 32.9 W Ger Norway 30.5 Neth. Japan 283 Norway Chile 25.3 Switz. Rate 30.4 23.1 19.4 17.2 16.0 15.2 13.5 12.5 11.2 10.5 10.3 Blot, 1982). Men in Scotland have the highest lung cancer death rate in the world, followed by men in the Netherlands, England, Wales and Northern Ireland. The lung cancer death rates are relatively low in Chile and in other Latin American coun- tries. They are also low in China, India, and most 20 40 60 80 100 Oral cavity A Cancers of the oral cavity generally refer to can- cers located in the mouth and throat. The highest death rates in the world—three times higher than death rates found anywhere else—are found among both men and women in Hong Kong. In all other countries, the death rate among men is three or more times higher than for women. The high death rates from oral cancers in Hong Kong are chiefly due to cancers of the nasopharynx, a cancer that is rare among North Americans and Europeans. High death rates from nasopharyngeal cancers have also been found in parts of China. Death rates from oral cancers are lowest in Japan and in the Netherlands. Death rates per 100,000 pop- ulation for oral cancer In 20 countries, 1976—77, age-ad- justed to U.S. standard 1970 population. From: World Health Statistics Annual 1979—80 as adapted by American Cancer Society 1983. Country Rate Male Country Rate Female H. K. 21. 2 H. K. 7.1 Swrtz 8.0 US 2.0 Hungary 7 6 Scotland 1.9 U 8.5.8 England 1.7 Poland 5.8 N Zea 1.7 Australia 53 N. Ire 1.6 Canada 5. 3 Canada 1.6 Austria 5. 2 Denmark 1.5 Scotland 4 4 Israel 1.5 Norway 4. 3 Australia 1.5 N. Zea. 4. 3 Hungary 1.4 England 3. 7 Sweden 1.4 W. Ger. 3 5 Swilz. 1.3 Sweden 3. 4 Poland 1.3 N. Ire. 3. 3 Norway 1.1 Chile 3. 3 Austria 0.9 Denmark 31 Chile 0.9 Israel 2.8 W. Ger. 0.9 Neth. 2.5 Neth. 0.8 Japan 2.2 Japan 0.8 —-———————-Ia_-o 100 13 Cancer Death Rates in 20 Countries Prostate Death rates per 100,000 pop- ulation for cancer of the prostate in 20 countries, 1976—77, age-adjusted to U.S. standard 1970 population. From World Hea/tn Statistics Annual 7979—80 as adapted by Amer/can Cancer SOC/Qty 7983 Country Sweden Norway Swnz VV Ger Hungary New N Zea Aosnana U S Aumna Denmaw Canada N lm England SeoHand Chde Israel Poland Japan H K Ram 323 298 285 241 239 238 238 229 223 219 214 270 199 t8? 172 170 111 10.2 36 29 Cancer of the prostate is one of the most common cancers among men. ()f the 20 countries listed below. the highest death rates from this cancer are seen among northwest European men: Swedes. Norwegians. Swiss. East Germans. Hungarians. and Dutch in the Netherlands. In the United States. cancer of the prostate is the second most common cause of cancer death among men. This cancer is rare in east Asia: the death rate among Japanese men is 3.6 per |()().()()() a year, and the death rate of 3.9 in Hong Kong is the lowest in the world. "-IIIIIIIIIIIIIII|||§ ('D 60 80 tOO O N O a. O Stomach The Japanese have the highest death rates in the world for stomach cancer: 70.2 deaths per l()0.000 population for men and 34.9 for women. Stomach cancer causes more deaths among the Japanese than all other types of cancer combined. Ameri- cans. in contrast. have stomach cancer death rates among the lowest in the world. The death rates for this cancer are relatively high for both men and women in Chile. Hungary. Austria. Germany. and Poland. They are significantly lower. for both sexes. in Israel. Canada. and Australia. The U.S. death rates from stomach cancer were high until the 1930s; they have since been declining steadily. Death rates per 100,000 pop- ulation for stomach cancer in 20 countries, 1976—77, age- adjusted to U.S. standard 1970 population. From World Health Statistics Annual 7979780 as adapted by Amcrrcan Cancer Society 1983 Country Japan Chile Hungary Poland Auslna W Ger Neth Scotland England Norway N Ire Sweden N Zea Denmark Israel Canada Australia Sw1tz U S H K Rate 70 2 64.9 47 4 44 6 38.5 34 4 28.2 25,4 25 1 23.8 23 4 2O 9 19.1 19 O 18 4 15.8 15 7 12 0 9 2 5 8 Male Country Japan Chile Hungary Austria W. Ger Poland N. Ire Sconand Neth Norway England Swntz Sweden Denmark Israel H K N Zea Australia Canada U 8 Rate 349 304 226 206 18.3 175 14.3 134 130 118 118 116 108 97 92 92 85 84 72 44 -..--IIIIIII|IIIII§ N 5 O l'\) O 4:. O C‘; C CD O O O 40 60 80100 l5 C (\J O Cancer Death Rates in 20 Countries Uterus Death rates per 100,000 popu- lation for cancer of the uterus in 20 countries, 1976—77, age-adjusted to U.S. standard 1970 population. From, World Health Stat/st/cs Annual 1979a80 as adapted by Amer/can Cancer Society 7983. Country Chile Hungary Austria Poland Denmark W Ger H K Japan Swuz Scotland England N Zea Norway Sweden Neth‘ U.S. N lre Canada Australia Israel Rate 223 196 163 157 152 121 121 11.8 112 105 104 102 99 96 93 8.9 84 83 79 58 Cancers of the uterine cervix and of the uterine corpus. or endometrium. differ in type. cause. and risk factors. Some countries. however, do not dis~ tinguish between the two in their reporting. thus limiting the usefulness of international com- parisons. Both types are included in the chart above. Among the countries represented. Chilean women have the highest death rate for the two can- cers combined‘ and Israeli women the lowest. Death rates for these two cancers are relatively low in the United States. Both the incidence and death rate for cervical cancer have been decreasing in the past several decades in the United States and in some other Western countries. but rising in others. III-IIIIIIIIIIIIIIII§ P. 0 O M O 4:. O 60 80 100 International Range of Cancer Incidence Although cancer occurs in every country in the world. there are wide geographic variations in incidence. Among men. for example. lip cancer occurs at a rate of 22.8 cases per 100.000 population in Newfoundland and 0.1 cases per 100.000 in Osaka. for a high—low ratio of 228. The incidence of cancer of the pros— tate is highest among black men living in Alameda County. California. and lowest among men in Shanghai. The high-low ratio is 125.3. Stomach can— cer occurs at a rate of 100.2 cases per 100.000 among Japanese men in Nagasaki and 5.7 cases per 100.000 among US. white men in Atlanta. for a high-low ratio of 17.6. Both the world‘s highest and lowest incidence rates of lung cancer in men are found in the United States: the highest incidence rate—107.2 cases per 100.000—is found among black men in New Orleans and the lowest—8.1 per 100.000—is found among American 1n- dians living in New Mexico. This latter group has the world‘s highest inci- dence rate of gallbladder cancer—7.7 cases per 100.000—compared with 0.5 cases per 100.000 among men in Bombay. Among women. the greatest world- wide variation in incidence is found for cancer of the nasopharynx. It occurs at a rate of 0.1 cases per 100.000 popu- lation in Trent. England. and at a rate 144 times greater. 14.4 cases per 100.000. among women in Hong Kong. The highest incidence rate of breast cancer in the world. 87.5 cases per 100.000. is found among Hawaiian women and the lowest rate. 8.9 cases per 100.000. is found among Japanese women in Osaka. The high—low ratio is 9.8. Cancer of the uterine corpus. or en- dometrium. occurs at a rate of 38.5 cases per 100.000 among white women in Alameda County. California. and at a rate of 1.0 cases per 100.000 among rural Japanese women in Fukuoka. a 38.5-fold difference. The world‘s highest incidence of melanoma is found in New South Wales. Australia. It occurs there at a rate of 19.1 cases per 100.000 among women and 17.2 cases per 100.000 among men. The lowest incidence of this cancer is found among Japanese men and women in Osaka. By comparing the worldwide inci— dence of cancer among men and wom- en. it can be seen that cancer is gener- ally less common among women. The table also shows that the United States has the highest incidence rates in the world of a number of cancers: breast and bladder cancers among women. prostate and kidney cancers among men. and cancers of the colon. pan- creas. and gallbladder and myeloid leukemias in both sexes. The US. incidence rates for stomach cancer among men. lung cancer among Amer»- ican Indian men. and cancer of the esophagus among women are the lowest in the world. The charts on the following two pages are based on data obtained by the International Agency for Research on Cancer (lARC). a part of the World Health Organization. [ARC was formed in1965 to collect and evaluate international data on cancer and to identify potential risk factors through epidemiologic and laboratory studies. International range of incidence for selected sites of cancer around 1976 by sex. Males High Low Ratio Site Population Rate Population Rate H/L Lip Canada. 228 Japan. Osaka 0.1 228.0 Newfoundland Tongue India. Bombay 10.2 Romania. County Clul 0.5 20.4 M0uth France. Bas Rhin. 130 Japan. Miyagi 0 5 26.0 Urban Oropharynx France. Bas Rhin. 134 Norway 0.3 44.7 Urban Nasopharynx Hong Kong 329 Japan. Mlyagi 0 3 109.7 Hypopharynx France. Bas Rhin. 11.0 Israel. All Jews 0.2 55.0 Rural Esophagus Shanghai 24.7 Hungary, Szabolcs. 1 1 22.5 Rural Stomach Japan. Nagasaki 100.2 U.S.. Atlanta. White 5.7 17.6 Colon U.S.. Connecticut 323 India. Poona 3 1 10.4 Rectum Canada. NW, Territory 226 Israel. Non»Jews 3,1 7,3 & Yukon Liver Hong Kong 344 Australia. New Soulh 0.6 57.3 Wales. Rural Gallbladder U.S.. New Mexrco. 7.7 India. Bombay 0.5 15.4 Amerindian Pancreas U.S.. Bay Area. Black 18.3 India. Bombay 2 0 9,2 Larynx Italy. Varese 16.0 UK, North Scotland 18 8.9 Lung & Bronchus U.S.. New Orleans. 107.2 U.S.. New Mexico. 81 13.2 Black Amerindian Melanoma Australia. New South 172 Japan. Osaka 0 2 86,0 Wales. Urban Prostate U.S.. Alameda. Black 100.2 Shanghai O 8 125.3 Testis Switzerland. Vaud. 105 Cuba 03 35 0 Rural Penis Jamaica. Kingston 57 US. Los Angeles. 0.2 28.5 White Bladder Switzerland. Geneva 30.2 India. Poona 24 12.6 Kidney. etc. U.S.. Hawaii. White 11.2 India, Bombay 1.3 8.6 Brain Australia. South 82 Japan. Miyagi 0.9 9,1 Thyroid gland U.S.. Hawaii. Chinese 7.8 India. Poona 0.4 19.5 Lymphosarcoma Switzerland. Geneva 8.5 Poland. Warsaw. Rural 1 2 7 1 Hodgkin‘s Switzerland. Vaud. 4.9 Japan. Miyagi 0.5 9.8 disease Urban Multiple myeloma U.S.. Bay Area. Black 8.4 India, Poona 0.6 14.0 Lymphatic Switzerland. Neuchatel 7.9 Japan. Fukuoka. 0.5 15.8 leukemia Urban Myeloid leukemia U.S.. Hawaii. Hawaiian 8.7 Romania. County Cluj 0.7 12.4 ‘Age—standardrzed to the Standard World Population (Waterhouse et al. 1982). I8 International range of incidence for selected sites of cancer around 1976 by sex. Females High Low Ratio Site Population Rate Population Rate H/L Lip Romania, County CIuj 2.3 UK, Birmingham 0.1 23.0 Tongue India, Bombay 4.1 Czechoslovakia, W. 02 20.5 Slovakia Mouth India, Bombay 5.8 Yugoslawa, Slovenia 02 29.0 Oropharynx US, Hawaii, White 27 Japan. Osaka 0.1 27.0 Nasopharynx Hong Kong 144 Trent. UK. 01 144.0 Hypopharynx India. Bombay 2.2 Canada, British 0.1 220 Columbia Esophagus India. Bombay 10.7 U.S., Utah 04 26.8 Stomach Japan, Nagasaki 51.0 Israel. Non-Jews 24 21,3 Colon 08., Bay Area, 27.4 India, Poona 2.8 9.8 Japanese Rectum SWItzerland, Neuchatel 13.4 Israel, Non-Jews 1.5 8.9 Liver Shanghai 91 Norway. Rural 0.4 22.8 Gallbladder US, New Mexico, 22.2 India. Bombay 07 31.7 Amerindian Pancreas US, New Mexico, 104 India. Bombay 09 11.6 Amerindian Larynx India, Bombay 2.6 Norway 02 13.0 Lung & Bronchus New ZeaIand, Maori 48.8 Spain, Navarra 2.6 18.8 Melanoma Australia, New South 19.1 Japan. Osaka 0.2 955 Wales, Rural Breast US, Hawaii. Hawaiian 875 Japan. Osaka. Rural 8,9 9.8 Cervix uteri Colombia. Cali 52.9 Israel, Non-Jews 2.1 25.2 Corpus uteri U.S., Alameda, White 38.5 Japan. Fukuoka, Rural 1.0 38.5 Ovary Israel, Jews born in 17 2 Japan, Osaka, Rural 2.1 8.2 Europe 8. America Bladder US, New Orleans, 6.5 Hungary. Szabolcs, 0.5 13.0 White Rural Kidney, etc. Canada, N.W. Territory 153 India. Poona 0.6 25.5 & Yukon Brain Poland, Warsaw City 66 Japan. Miyagi. Rural 0.6 11.0 US, Hawaii Thyroid gland Hawaiian 17.6 Poland, Warsaw, Rural 0.7 25.1 Lymphosarcoma US, Hawaii. Hawaiian 6.3 Poland. KatOWIce 05 12.6 Hodgkin's Switzerland, Vaud. 43 Japan. Osaka 0.3 14.3 disease Rural Multiple myeloma U 8., Hawaii, Hawaiian 5.9 Poland, Katowice 0.4 14.8 Lymphatic U.S.. New Mexico, 3 3 Japan, Fukuoka 0.4 8.3 leukemia Other White Myeloid leukemia New Zealand, Maori 5 4 Hungary. Szabolcs O 8 6,8 ‘Age-standardized to the Standard World Population (Waterhouse et a1, 1982). U.S. Cancer Incidence A§;5;ecific, All Sites, by Age and Race Average annual age-specific cancer incidence per 100,000 U.S. population, all sites combined, by race and sex, 1973—77. From: SE ER. Death rate per 100.000 Cancer is chiefly a disease of middle and old age. rare in children and young adults. More than half of all cases of cancer are diagnosed after age 65. Up to age 50. the incidence of cancer is higher in women. After age 60. there is a dramatic in- crease in cancer incidence among men. 3500 Age 0 10 20 — White Males _ White Females — Black Males Black Females 30 4O 50 60 7O 80 U.S. Cancer Incidence By Age Group for All Races and Both Sexes Incidence of cancer per 100,000 U.S. population, all slles combined, all races and both sexes, by age group, 1973-79. From: SE ER, Data in this chart are set up on a logarithmic scale to display percentage changes in cancer incidence in five different age groups; the log scale yields a truer picture of time changes than do arithmetic scales. The greatest changes in incidence in the years covered were among adults aged 65 to 74. Lesser increases are seen in adults aged 45 to 64. while incidence has remained stable for ages 15 to 44. Incidence has decreased for children under age 14 (the 1975 dip for this age group may be due to changes in the population surveyed by SEER). Death rate per 100,000 LOCO I 44F l l llllT 100 l l lllll l lllllll T 1 _ 1974 1973 1975 1976 1977 1978 1979 U.S. Cancer Incidence 1 0 Most Common Cancers by Sex Among Whites, Blacks, and Hispanics, 1 973—77 Cancer of the lung is the most common type of cancer among black and white men with an inci- dence of l l0 cases per |()0.000 population among blacks and 76 per 100,000 among whites. Hispanic men are at less risk for this cancer. Cancer of the prostate is the most common can- cer among Hispanic men with an incidence ap- proaching 59 cases per [00,000. Among black men. the incidence of cancer ofthe prostate is almost as high as lung cancer, near 110 per 100,000. Colorec- tal cancers are the third most common cancers among both black and white men, and the fourth among Hispanic men. Stomach cancer is the third most common can- cer among Hispanic men. the fourth among black men. and the seventh among white men. White men are at far greater risk for bladder cancer than either black or Hispanic men. Breast cancer is the most common cancer among women of all three groups. It occurs most frequently among white women. with an incidence of 85 per 100,000 population. compared with 75 per l00.000 in black women and 48 per 100,000 in His- panic women. Colorectal cancers are the second most common among all three groups of women. Cancer of the uterine corpus is the third most frequent cancer among white women; cervical can- cer is third among black and Hispanic women. Lung cancer is fourth among all three groups of women. ’7') White mates Whtte tomatcs Lung _ Breast — Prostate — Coton-Bectum — CotoniFtectum _ Corpus — Bladder - Lung - Lymphomas - Ovary - Leukemras - Cervrx - Stomach - Lymphomas - Pancreas - Pancreas I Krdney - Leukemlas I Larynx I Bladder I Btack mates Btack tomalos Lung _ Breast — Prostate _ CO’OWRBCTUW _ Colon-Rectum — Cervrx _ Stomach - Lung - Pancreas - Corpus - Esoohagus - Pancreas - Btaddcr - Stomach - Larynx - OVHW - Leukemtas - Loukemras I Mull Myeloma - Mult Myeloma I Hrsparnt: mates Htspahrt: tomates Prostate ‘ Breast _____ _~fi___j Lung CotoniRectum 7‘_ J' Stomach : Cervrx ‘1» J‘ Cotonfleclum Lung J ,.J Pancreas . Stomach V___] Lymphomas V Pancreas _‘ Btadder ‘ Ovary JJ Leukemras Corpus J; Kidney ' GaHbtadder Wfl Bram Lymphomas __J 0 20 4O 60 80 100 120 0 20 40 60 80 100 120 tncrdentto rate per 100 000 (agvxadlustmt to 1970 U 8 standard) Incidence per 100,000 U.S. population of 10 most com- mon cancers among whites, blacks, and Hispanics, by sex, 1973—77, age-adjusted to 1970. From. SE E R U.S. Cancer Incidence Changing Patterns by Site Among Whites and Blacks, Men and Women The incidence of cancers by site for men and wom- en, white and black, for five different time periods appear in these two charts. The data for 1969 and 197] came from the Third National Cancer Survey, covering a population of 21 .003,451 from nine ma- jor areas in the United States. The data for 1973, 1975. and 1977 came from SEER, which then covered 11 geographic areas representing about 10 percent of the U.S. popula- tion. The populations are not strictly comparable. but they can be used to give an idea of changes in cancer incidence, by site. over a period oftime. Incidence for whites, both sexes, per 100,000 U.S. popu- lation by year and site, all ages combined, age-adjusted to 1970. Data for 1969 and 1971 from TNCS. Data for 197371977 from SEER 24 Incidence per 100,000 Cancer Site Sex 1969 1971 1973 1975 1977 Bladder M 23.8 23.4 25.5 25.8 26.3 F 6.3 6.3 6.1 6.9 7.2 Breast F 73.9 75.1 81.0 86.2 82.7 Colon M 34.5 32.4 34.2 35.5 38.5 F 30.6 28.6 29.7 30.6 32.1 Kidney M 9.0 8.2 9.4 9.0 9 4 F 4.3 3.8 4.4 4.0 4.6 Lung M 70.6 70.0 72.3 76.4 79.4 F 13.3 15.5 17.7 21,8 24.5 Melanoma M 4.4 4.7 5.8 6.4 7.6 F 4.1 4.8 5.1 6.0 6.7 Ovary F 14.9 13.6 14.2 14.2 13.7 Pancreas M 12.1 12.3 12,7 12,5 11.6 F 7 5 7.0 7 5 7 2 7.5 Prostate M 59.0 56.7 61,0 64.8 70.4 Rectum M 17.5 18.1 18 8 18.3 191 F 11.1 10.6 113 12.0 11.2 Stomach M 15.4 13.4 13.8 12.7 11.6 F 7.1 6 3 6 1 5.4 5.2 Total Leukemia M 13.2 12.2 13.2 12.5 11.7 F 8.0 7.2 7.8 7.3 7.6 Uterine Cervix F 16.0 14.3 12.6 10.7 9.5 Uterine Corpus F 22.6 24.6 29.0 32.4 28.5 I Melanoma incidence has almost doubled for white men and women. I Breast cancer incidence has increased sharply among women of both races but is more frequent among white women. I The incidence of cancer of the uterine cervix has decreased markedly among black and white women but its incidence remains more than two times higher in black women. I Cancer of the uterine corpus. or endometrial cancer. has increased in incidence among women of both races. Its incidence is markedly higher among white women. Incidence for blacks, both Incidence per 100,000 sexes, per 100,000 U.S. popu- Cancer Site Sex 1969 1971 1973 1975 1977 Iation by year and site, all _ , Bladder M 13.2 9.9 10.6 13.0 16.4 ages combined, age-adjusted F 4.2 5.0 3.8 5.2 5.5 to 1970. Data {0, 1969 and 1977 from Breast F 62.1 54.9 66,9 75.9 71.5 TNCS- Data 10’ 7973-7977 Colon M 29.0 26.0 31.9 31.8 42.9 from SEER F 28.1 29.0 29.5 32.4 29,8 Kidney M 7.4 7,2 8.7 8.1 9.4 F 3.8 4.0 4 4 4.0 4.5 Lung M 78.6 102.8 108.4 107.4 112.7 F 13.5 15.0 21.7 21.7 28.4 Melanoma M 0.5 0.8 0.4 0.8 0.3 F 0,8 0.8 0.7 0.7 0.6 Ovary F 9.5 10.9 9.9 10.1 8.6 Pancreas M 15.1 16.8 15.8 15.3 17.7 F 8.4 11.1 11.9 12.0 12.1 Prostate M 99.9 94.1 107 6 1119 115.8 Rectum M 14.4 14.4 11.2 13.5 14.1 F 9.7 8.7 11.9 11.2 10.7 Stomach M 22.1 24.2 27.2 21.7 19.8 F 8.0 10.8 9.7 10.4 9.6 Total Leukemia M 11.6 9.8 12.5 11.4 10.0 F 6.2 5.5 8 3 6.4 5.6 Uterine Cervix F 35.2 32.8 30.3 27 2 22.2 Uterine Corpus F 11.3 14.5 14,7 17.2 16.8 [J 'JI Ovarian cancer incidence has dropped slightly among black and white women. Incidence of cancer of the prostate has increased among men of both races but most markedly among black men. Bladder cancer incidence has increased for all four groups. It is higher among whites than among blacks. Leukemia incidence has decreased for all four groups. The incidence of kidney cancer has increased slightly among all four groups. Stomach cancer incidence has declined among black and white men and among white women. It has increased among black women. Colon cancer incidence has increased sharply among black men since 1969. It has increased gradually among white men. Colon cancer inci- dence has increased slightly among white and black women. The incidence of cancer ofthe rectum has in- creased among white men and women and black women: it has decreased slightly for black men. Pancreatic cancer incidence has decreased slightly among white men and remained constant among white women. It has increased signifi- cantly among black men and women. Lung cancer has more than doubled among black women in the 8 years covered by these data. and has almost doubled among white women. Marked increases are also seen for black and white men. U.S. Cancer Death Rates Changing Cancer Patterns by Age Group for All Races and Both Sexes, All Sites Combined Cancer death rates, by age group, per 100,000 u.s. popu- lation for all races and both sexes, all sites combined, 1 965—1 979. From: NCHS, USBC Data reflected in this chart are displayed on a logarithmic scale to show percentage changes in cancer death rates in five different age groups. The death rates for children and younger adults have both declined markedly since 1964. and ap- pear to be leveling off. There have been slight declines in death rates among persons aged 45 to 64 in the past several years. while death rates for persons aged 65 to 74 have remained almost constant since I964. Death rates for persons 75 and over have risen slightly since 1970. Death rate per 100000 I 1,000 Ages 75+ Ages 65—74 llllllll Ages 45—64 I 100 Illlllll l Ages 15—44 * Ages 0—14 llllllll 1 _ l l l l I I l 1966 1968 1970 1972 1974 1976 1978 1980 27 U.S. Cancer Death Rates Changing Patterns for Eight Major Cancers in All U.S. Men Death rates for males, per 100,000 U.S. male population, for eight cancer sites, 1930—1977. Age-adjusted to 1970. From‘ NCHS‘ USBC Lung Colon 8. Rectum Prostate Pancreas Stomach Leukemia Bladder Esophagus The lung cancer death rate for all U.S. men has climbed fourteenfold in the past 50 years. from 5 deaths per 100,000 population in 1930 to almost 70 deaths per 100,000 in the late 1970s. Stomach cancer was the leading cause of cancer deaths among men from I930 until 1947. Deaths from this cancer, however. have decreased markedly since 1930. from 37 deaths to 9 deaths per 100,000 population. Deaths from colorectal cancers and cancers of the prostate both peaked in the late l940s. then dropped and leveled off. Deaths from cancer of the pancreas and from leukemia rose slightly from the 1930s into the l960s and then leveled off. Esophageal and bladder cancer deaths have re- mained almost constant since 1930. 7O 60 50 40 30 20 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 U.S. Cancer Death Rates Changing Patterns for Eight Major Cancers in All U.S. Women Death rates for females, per 100,000 U.S. female popula- tion, for 8 cancer sites, 1930-1977. Age-adjusted to 1970. From: NCHS, USBCI . Breast ‘-——— Colon & Rectum I l— Lung Uterus Ovary I! —— Pancreas Leukemia —— Stomach Until about I945. cancer ofthe uterus was the ma- jor cause of cancer death among all U.S. women. but deaths from this cancer have been declining steadily since 1930. Breast cancer deaths have risen only slightly since 1930. but breast cancer has been the major cause of cancer deaths among all U.S. women since about 1945. Lung cancer deaths, meanwhile. have been climbing sharply since the early 1960s. and are ex- pected to overtake breast cancer deaths sometime in 1984 or 1985. Stomach cancer deaths have been declining steadily since 1930. from 28 to less than 5 deaths per 100.000 population. Cancer of the uterus‘ once the major cause of cancer deaths among U.S. women. now accounts for less than 10 deaths per 100.000. The death rates from leukemia and ovarian can- cer both rose slightly from I930 to 1955 but have since leveled off. Pancreatic cancer deaths rose slightly over the period and are still rising slightly. The death rate for this cancer is still less than 8 per 100,000 each yean Deaths from colorectal cancers rose in the l930s and l940s but have been declining since I945. 40 _\ 30 / 20 v ‘ 10 l I I I l l I I I I I 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 29 U.S. Five-year Survival Rates Changes for 19 Sites Among Whites The overall. 5-year relative survival rate for cancer patients in the United States is now 48 percent or greater. These data show the changes in survival rates for all whites in this country for 19 major cancers. There have been some gains in survival rates for almost all cancers in the 20 years spanned by these data. The most notable gains have been seen for endo— metrial. cervical and breast cancers among wom- en. testicular and prostatic cancers among men. and for Hodgkin‘s disease. melanoma skin cancer. and bladder cancer among both men and women. Survival for colorectal cancer patients has im- proved over the 20-year period but is still less than 50 percent. Survival rates for cancers of the lung. stomach. pancreas. and esophagus remain low. Five-year relative survival rates for selected sites for white cancer patients. 30 ‘Data for 196(k63 and 1970—73 are from three hospital registries and 1960—63' 1970—73~ 1973—79‘ Bladder 53 61 71 Brain 18 20 20 Breast (females) 63 68 72 Cervix 58 64 66 Colon 43 49 48 Endometrium (corpus) 73 81 87 Esophagus 4 4 4 Hodgkin's disease 40 67 68 Kidney 37 46 48 Leukemia 14 22 28 Lung 8 10 11 Melanoma of the skin 60 68 76 Non—Hodgkin‘s lymphoma 31 41 48 Ovary 32 36 34 Pancreas 1 2 2 Prostate 50 63 64 Rectum 38 45 46 Stomach 11 13 g 13 Testicular cancer 63 72 l 80 one State registry, and appear in Cancer Patient Survival Expltile ice. 1980 Data for 1973—79 are from SEER. and represent 10 perr‘enl of the U S population. Thus. the earlier data and the SEER data am not strict- ly comparable but each set represents the best available data .or the time period covered. h 0.5. Five-year Survival Rates Changes for 1 9 Sites Among Blacks Substantial gains in 5-year survival rates can be seen for endometrial, cervical and breast cancers among women and for cancer of the prostate among men. Some gains in survival are also apparent for bladder cancer, kidney cancer and colorectal can— cers in both sexes. The outlook for stomach cancer and for cancers of the lung. pancreas. ovary and esophagus remains poor. Although white patients survive longer than black patients for more than half of these sites, the survival rates are almost equal for cancers of the stomach. lung and esophagus. and for Hodgkin’s disease and non—Hodgkin’s lymphoma. Black pa- tients with cancers of the pancreas, ovary. kidney. and brain survive slightly longer than whites. The differences in survival rates is a subject of concern and is under intense study at the National Cancer Institute. Five-year relative survival rates for selected sites for black cancer patients. 1960—63' 1970—73‘ 197377? Bladder 24 36 43 Brain - 19 19 21 Breast (females) 46 51 60 Cervix 47 61 61 Colon 34 37 44 Endometrium (corpus) 31 44 54 Esophagus 1 4 3 Hodgkin's disease ** *‘ 66 Kidney 38 44 49 Leukemia ** ** 24 Lung 5 7 9 Melanoma of the skin ** ** " Non—Hodgkin's lymphoma " " 43 Ovary 32 32 35 Pancreas 1 2 4 Prostate 35 55 54 Rectum 27 3O 35 Stomach 8 13 14 Testicular cancer " " 62 ‘Data lor 1960763 and 1970—73 are from two hospital registries and ap- pear in Cancer Patient Survwal Experience, 1980 The later data are from SEER. and represent 10 percent of the U 8, population. Thus. the earlier data and the SEER data are not strictly comparable, but each set represents the best available data for the time periods covered ‘Rates could not be calculated because the number of cases was too small n 31 U.S. Five-year Relative Survival Rates Changes for white children under age 1 5 Dramatic improvements in 5-year survival rates have been achieved in the past 20 years for white children under age 15. (Black children were omit- ’ ted from this analysis because ofthe small num- bers of patients.) Major gains in survival are seen for acute lym— phocytic leukemia, or ALL. the most common leu- kemia among children. The outlook for acute gran— ‘ ulocytic leukemia remains grave. The outlook for children with cancers of the brain and central nervous system, Wilms‘ tumor. neuroblastoma. bone cancers. and Hodgkin's dis— ease has improved over 1960. Survival rates for children with non-Hodgkin‘s lymphomas have improved but are still only 38 percent. at p Five-year relative survival rates for cancer in white chil- dren under age 15. 1960—63’ 1970—73“ 1973—79‘ ' Acute lymphocytrc leukemia 4 34 56 Acute granulocytrc leukemia 3 5 19 Bone 20 30 45 Brain and central 35 45 51 nervous system i Hodgkin's disease 52 90 83“ Neuroblastoma 25 4O 46 ' Non—Hodgkin's lymphoma 18 26 38 Wilms' tumor 57 69 72 ‘ “Data for 196(k63 and 1970773 are from three hospital registries and one State registry. and appear In Cancer Patient Survrval Experience. 1980 Data for 1973—79 are from SEER. Wthh represent 10 percent of L the U 8 population The earlier data and the SEER data are not strictly comparable but each set represents the best available data for the L time periods covered "An apparent decrease In 5»year relative survrval for children With Hodgkin's disease appears because of the change In data bases The latest data can be applied to the U 8 population The t970~73 data rellect the inclusron of results from an outstanding research center. 4 one at the tour registries used lor the earlier data U.S. Costs of Cancer Major Medical Expenditures, 1980 Americans spent about $2|9.4 billion on health care in 1980. This figure includes costs for diag- nosis. hospitalization. physicians. drugs. long—term care. and rehabilitation. Costs for the four major causes of death in this country—heart disease, cancer. strokes and inju- ries—together accounted for about a quarter of the total health care bill. Accidents and injuries, the leading cause of death among children and the fourth leading cause of death among adults. cost Americans $19.2 bil- lion, or 9 percent of the total bill. Heart disease, the leading cause of death among adults. cost $14.5 billion or 7 percent. and cancer. the second leading cause of death among adults. cost $13.] billion. or 6 percent. These costs are derived from the records of health insurance companies. hospitals, physicians. and state and federal health care plans. The data were compiled by the National Center for Health Statistics ofthe US. Public Health Service. Medical care expenditures for cancer, heart disease, stroke, and injuries, 1980 (in billions). From' NCHS, /——- 7% $219.4 $192 $145 6 0/0 $131 2% 9% All diseases Cancer Heart disease Stroke Injuries 33 U.S. Costs of Cancer By Cancer Sites and Age Group Medical care expenditures (in billions) for cancers and benign tumors, 1980; total $13.1 billion, excluding nurs- ing home care. From' NCHS Medical care for benrgn tumors In- cludes care for nonmalignant skin and brain tumors and rn srtu car- crnoma ot the cervtx, and blOpSleS tor benign breast disease, - Male - 65 - Male '65 - Female - 65 _Female -65 Digestrve organs Lung. bronchus. and trachea Skrn. bone. and connectrve trssue Breast Female genital organs Other genitourinary orgrans Lymph and blood systems Other and unspecrlred Sites Benrgn tumors’ 34 This chart shows. by cancer type. the $13.1 billion spent on cancer in 1980. Costs for each cancer type are also figured for by sex and age—under 65 and over 65. To some extent. these costs reflect the incidence of specific cancers. Breast cancer. for example. is the major cause of cancer among women. and ac- counts for a major portion of the health care dollar spent on cancer. Costs of care for benign tumors include diag- nosis and treatment of non—cancerous skin tumors in both men and women. and of in Si!!! carcinoma of the cervix. The cost figure also includes biopsies for benign breast disease. Care for colorectal cancer. the second most fre- quent cancer among women and the third among men. is included in the figure for cancers ofthe di— gestive organs and peritoneum. Care for lung cancer. a major cancer in both men and women. accounts for most of the care costs for cancers of the lung. bronchus and trachea. The figure for female genital organs includes care for cancers of the uterine corpus. or endome- trium. and ovarian cancer. That for other genitouri- nary organs includes cancers ofthe bladder and kidney in both men and women. and for cancer of the prostate and testicular cancer among men. The lymph and blood systems category includes leukemia and lymphomas. Percent ~ 65 '65 48% 52% $1 1 26% 74% $0 9 , 55% 45% $1 0 31:21 74% 26% $0 6 - 61% 39% $0 4 4;». 66% 34% $0 5 —6£§Wkez; 66% 34% $1 2 —r—4Ax 72% 28% $0 9 fi 32% 68% so 9 ' 40% 60% $0 2 — 61% 39% $0 5 5K: :1 61% 39% $0 5 ‘1 69% 31% $0 7 72% 28% $06 ‘ sigh-g. 87% 13% $2 2 » u.s. Costs of Cancer Earnings Lost Due to Cancer in 1 977 The value of lost earnings due to cancer and benign tumors was estimated to be $26.4 billion in 1977. “Lost earnings" are those lost because of pre- mature death from cancer. They are computed by multiplying the number of cancer deaths at various age levels by adjusted projections of lifetime earn- mgs. Even though 60 percent of 1977 cancer deaths occurred in persons 65 and older, that age group accounted for only 11 percent of lost earnings. Those aged 45 to 64 accounted for 34 percent of deaths and for 62 percent of lost earnings. Persons under age 45 at time of death were responsible for 6 percent of cancer deaths and for 27 percent of lost earnings. Lung cancer deaths accounted for the largest single portion of lost earnings: almost $6 billion. Cancers of the digestive organs include colorectal cancer and cancers of the stomach, esophagus. and pancreas. Earnings lost (in billions) due to cancers and benign tu- mors, 1977; total: $26.4 billion. From: NCHS. All cancer and benign tumors - Morbidity Digestive organs $5 2 Lung bronchus and trachea $6 0 Skin $0 6 Breast $2 7 Female genital organs $1.6 Other genitourinary organs $14 Lymph and blood systems $2 8 Other and unspecmed sites $3 7 $0.4 Benign tumors - Mortality 35 ‘ IIIIII'IIIIIIIIIIIII'II. p r , Risks Risk Factors Air and Water 38 As a nation we have learned that the air around us may become contaminated with industrial and au- tomotive emissions. and that this may happen to our seas. lakes. rivers. and groundwater supplies as well. Americans are becoming increasingly con- cerned with this issue, but there is still very little conclusive information about the effects of such pollution on the incidence of cancer. Each of us. at various times. may be exposed to potential carcinogens in the air we breathe and the water we drink. Though the number of such sub- stances is large. the levels are generally small— much smaller. for example. than the levels found in some workplaces. ()n the other hand. carcinogens in air and water may be harder to avoid than those found in specific workplace locations. It has been estimated that pollution accounts for at most I to 5 percent of all cancer deaths (Doll and Peto. 1981). Air Pollution There is little evidence to date that ambient air— the circulating air around us—poses serious cancer risks. The ambient air in specific areas may con- tain industrial plant emissions. automobile exhaust. and other pollutants linked to cancer. but in most places. the air we breathe does not contain high levels of carcinogens. In fact. the strongest link between air pollution and cancer is found among smokers. who inhale particulates in their tobacco smoke. Much of what we know about the health effects of large amounts of pollutants in the air comes from workplace studies ofcoal gas. tar. pitch. and coke oven emissions. Certain workers may be at some risk but workplace exposures are often l() to l()()() times the levels found in ambient air. Asbestos is a known carcinogen found in the air around some workplaces. Levels vary. but are highest near asbestos mines. mills. waste dumps. and manufacturing plants. Asbestos is the only known cause of mesothelioma. a form of cancer that affects the membrane lining of the chest and abdominal cavities. It is rare in the general popula- tion. but found frequently among workers exposed to high levels of airborne asbestos. Asbestos can also cause lung cancer. Studies of shipbuilders in coastal areas of Georgia. Florida. and Virginia. many of whom were exposed to asbestos insula- tion. showed an increased risk of lung cancer in all areas. and of mesothelioma in Virginia. particularly among employees who worked during World War ll (Blot and Fraumeni. l981). There may also be risks of exposure during demolition of buildings that contain asbestos used for fireproofing in walls and ceilings. There has been concern. for example. about children‘s exposure to asbestos in older school buildings. It is sometimes safer to cover interior asbestos than to try to remove it. Benzo(u)pyrene. or B(u)P. is a combustion prod- uct. B(a)P levels in ambient air in this country have decreased since the 19305 and l940s when oil and natural gas—which burn more cleanly—re- placed coal for home heating (Shy and Struba. 1982). A study of roofers exposed to high levels of B(u)P in hot pitch showed a higher incidence of lung cancer among them than in the general popu— lation (Hammond et al. 1976). Lung cancer incidence is higher in cities than in the country, but this is not necessarily due to ur- ban air pollution. Workplace exposures and other lifestyle factors ofcity dwellers. like cigarette smoking. may be more important (Buell. I967: Cederlof. 1975). The “urban factor“ is much weak- er than cigarette smoking on lung cancer inci- dence. Increased death rates from chronic. non-malig- nant lung diseases like bronchitis. particularly among the elderly. have been linked with “brown- outs." smogs. and other episodes of serious air pol— lution. Sulfur dioxide from older power plants and smelters is the main component of acid rain. It can also aggravate chronic lung disease. Few cancer studies have looked at any link with acid rain. but it produces other effects like damage to trees and to aquatic life in lakes and rivers. Exposure to radioactive emissions from radon in uranium mines has been shown to be responsible for the increased risk of lung cancer in miners. Re- cent concern has focused on radon exposures in homes. particularly those that have been made air- tight by efficient insulation. The radon can enter the home from soil. water. or building materials. Some researchers estimate that as many as 10.000 39 40 cancers a year may result in this country from in- door radon pollution. Cigarette smoke is another form of indoor pollution that has been linked to in- creased lung cancer. but the extent of the hazards from both radon and passive smoking is not now clear. Water Pollution Drinking water contains complex mixtures of known and suspected carcinogens including asbestos. metals. radioactive substances. and in~ dustrial chemicals. Even the process oftreating water may create small quantities of chemicals linked to cancer. but the levels are so small that there is probably a low risk. if any. associated with most drinking water supplies. Trihalomethanes. or THMs, can be formed when chlorine used to purify drinking water reacts with organic compounds in water. At levels normally found in chlorinated city water supplies. there is some suspicion that THMs may increase the risk of gastrointestinal and urinary tract cancers (Crump and Guess. I982). THMs are also used as indica- tors of more hazardous compounds that are diffi— cult to measure directly. To reduce the levels of THMs. water is often filtered so that less chlorine is needed to purify it. A few other chemicals that have been found in drinking water from a small number of supplies are ‘ known to cause cancer in humans (Harris et al. 1977). Vinyl chloride. for example. may be intro- duced into drinking water from industrial plants. or. in very small amounts, by seepage from poly— vinyl chloride piping used in some water distribu- tion systems. Benzene and bis(2-chloroethyl)ether are other carcinogens that are occasionally found in drinking water (Kraybill et al. 1980). Nitrates are seldom removed during the water treatment process. Nitrates themselves do not cause cancer. but they can be transformed in the body into nitrosamines, which are powerful car- cinogens. Although some studies have found a link between high levels of nitrates in drinking water and gastric cancer (Cuello et al. I976: Geleprin et al. [976). there is no indication that the low levels sometimes found in drinking water pose a risk of cancer L l Asbestos fibers are widely distributed in water supplies in this country. with higher levels often found near cities and industrial centers. But stud— ies have not shown consistently that asbestos in drinking water affects cancer risk (Shy and Struba. 1982). The trace metals arsenic. chromium. and nickel are found in drinking water in varying amounts. They may come from industrial plants. mines. seepage from soil or piping. by mineralization from rocks. or from water treatment processes. High levels of arsenic in drinking water in some other countries have been linked with skin cancer (Tseng et al. 1968; Tseng. 1976) but the much lower con- centrations in U.S. drinking water have not been linked with cancer in humans. Radioactive substances may be found in water depending on local rock type and on the use and disposal of radioactive compounds by nearby in- dustries. hospitals, and nuclear power plants. Ra- dioactive strontium and radium. sometimes found in some water supplies. can accumulate in bone tissue but even the cumulative dose from radium would result in so few fatal bone cancers that they would probably not be detected in epidemiologic studies. Naturally occurring radon gas is found dis- solved in water in some parts ofthe United States. Ingestion of radon in water does not pose much of a hazard, because of its low concentration. Radon can be released. though. from water into household air via showers and washing machines. Studies are trying now to evaluate the risk of such exposures. Water that percolates down below the earth‘s sur- face. known as groundwater. is often the source of spring and well water. Aquifers—rock formations that hold water in underground “lakes"—may also be sources for springs and wells. Contamination of these water sources with pesticides. industrial sol- vents. and other industrial chemicals such as poly- chlorinated biphenyls. or PCBs. can become a serious problem. Disposing of such wastes in lagoons or landfills can contribute to groundwater contamination. Burying hazardous wastes on land is the most common method of disposal in this country be- cause it is inexpensive. Disposal sites can leak. however. contaminating groundwater with toxic 4| Risk Factors substances. Contamination of groundwater could become more widespread as older disposal sites begin to leak. but cleanups and better disposal methods can prevent contamination. Environmen- tal Protection Agency regulations now require that hazardous waste disposal be completed in safer ways than in the past. The major United States water supplies are con- tinually monitored for carcinogens under Environ- mental Protection Agency guidelines. National Cancer Institute scientists have been studying the possible link between cancer and drinking water quality (Cantor et al. 1978). Except for some drink- ing water supplies that contain unusually high lev- els of carcinogens. the evidence to date suggests that our drinking water now poses little cancer risk to us. Alcohol Consumption of alcoholic beverages increases the risk of cancer. particularly ofthe mouth. pharynx. larynx. and esophagus (Tuyns. I982). Ethanol. by itself. does not cause cancer in animals. but very few people drink pure ethanol. They consume drinks flavored with congeners—chemical com- pounds or contaminants produced during fermenta- tion—and many people who drink also smoke. [I has therefore been difficult to evaluate the role of drinking alone. but there is no doubt that drinking and smoking together have a synergistic. or com- bined. effect. and do contribute to the high inci- dence of some cancers. This combined effect of drinking and smoking was shown clearly in a case-control study (Roth- man and Keller. 1972) of patients with cancers of the mouth and pharynx. Their risk increased with the amount of alcohol consumed and number of cigarettes smoked. Even among the nonsmokers in this study. though. an effect of alcohol could be seen. Risk for developing cancer of the larynx also increases with alcohol consumption (Wynder et al. 1976). although tobacco is a more powerful deter- minant. Esophageal cancer has been closely linked to drinking in a number of studies. Studies of black men in Washington. DC. (Pottern. Ziegler et al. l98l). where esophageal cancer rates are the high- est in the United States. showed that heavy drink- ing was chiefly responsible for the increased risk of esophageal cancer. In parts of China and Africa. where esophageal cancer incidence is also high. there is some link with “home breWS"—”sztmshu." a Chinese rice wine combined with grain alcohol (De Jong et al. I974). and in the Sudan. "aragi." a distillate made from dates (Malik et al. 1976). In Puerto Rico. a case-control study (Martinez. 1970) showed that alcohol and tobacco were the major factors associated with esophageal cancer. with alcohol having the greatest effect. And studies in France (Tuyns et al. I977 and I979) showed syn- ergistic effects of alcohol and cigarette smoking in increasing the risk of esophageal cancer. with local wines and ciders particularly implicated. ()ther cancers are also linked to alcohol con- sumption. although the evidence is not always clear cut. An increased incidence of liver cancer. for example. has been seen among patients with alcoholic cirrhosis. And a study done in Denmark (Jensen. 1979) reported a relative risk of LS for primary liver cancer among more than 14.000 brewery workers who were entitled to drink a beer “ration" each working day. But another study of brewery workers in Dublin (Dean et al. 1979) who consumed above-average amounts of stout failed to show a link with increased risk of liver cancer. About a fifth of the workers, who worked in the brewhouse. did have higher risks of cancer of the rectum. It is not now clear if one type of alcohol is riskier than another. chiefly because most ofthe popula- tions studied have drunk one type of alcohol only—wine. applejack. beer. grain alcohol. The Washington. DC. study (Pottern et al. op cit) did show that men who drank equivalent amounts (in terms of ethanol content) of hard liquor. or of hard liquor and beer. had a greater risk of esophageal cancer than did those who drank only wine and beer Since ethanol does not cause cancer in lab ani- mals. it is suspected that alcohol acts as a cocar- 43 Risk Factors cinogen. enhancing the effects of other cancer— causing agents like tobacco. It is also hypothesized that poor nutrition may contribute to the effects of alcohol by failing to provide protective nutrients. Diet 44 It has become clear that what we eat—or don‘t eat—for breakfast. lunch. and dinner may have profound effects on our chances of developing can- cer. One estimate (Doll and Peto. [981) is that the death rate from all cancers in this country might be reduced more than a third by practical changes in our diet. These are some of the diet components that are believed to affect the cancer process: Carotenoids and vitamin A Carotenoids are found in carrots. sweet potatoes. peaches. cantaloupes. and other yellow-orange fruits and vegetables. and in some dark green vege- tables like broccoli. kale. and spinach. Beta-car- otene and some of the other carotenoids are provitamins. or precursors. of vitamin A. When ingested. some of them are converted to vitamin A in the body. Preformed vitamin A. or retinol. occurs only in foods of animal origin—chiefly whole milk. Cheese. butter. egg yolks. and liver. Preformed vitamin A. in the form of retinyl acetate or retinyl palmitate. is also present in many multiple-vitamin pills and in some fortified foods. In many cultures. carotenoids from fruits and vegetables are the source of most of the vitamin A. and particular plant foods often account for most of them. Dark green. leafy vegetables are the main source of beta—carotene among some Chinese. for example. as are carrots in this country. a red palm oil used for cooking in West Africa. and yellow- green vegetables in Japan (Peto and Doll. l98l ). Inverse associations have been found. in a number of studies. between dietary vitamin A or carotenoids and the incidence ofa number of can- cers—those of the oral cavity and pharynx. the larynx. and lung (Doll and Peto. l98l: Shekelle et al. l98l). These findings suggest that either or both of these substances may protect against cancer. A study of esophageal cancer has suggested that def- ciencies of a number of micronutrients may in- crease risk for that cancer (Ziegler et al. 1981). To examine the effects of adding beta-carotene to the diet. the National Cancer Institute (NCI) began a study in 1982 of physicians enrolled in a National Heart. Lung. and Blood Institute (NHLBI) study. The NHLBI portion of the study is looking to see if regular aspirin use protects against stroke and heart attack; the NCI portion is looking at how beta-carotene affects cancer devel- opment. Other case—control studies are looking to see if fruits and vegetables. beta-carotene. or other factors are most strongly associated with de- creased cancer risk. and to examine which cancers are or are not influenced by diet factors. Vitamin C Some studies show an inverse association between fresh fruit and vegetable consumption and some cancers. including stomach cancer (Kolonel et al. 1981). though no studies have implicated vitamin C deficiency in the etiology of this cancer. Animal studies show that ascorbic acid—vitamin C—can inhibit the formation of carcinogenic N-nitroso compounds from ingested nitrates. Fats Surveys of various populations in this country and in others have suggested a link between dietary fat intake and some cancers. particularly those of the breast. colon. endometrium. and prostate gland. Animal studies tend to confirm these associations. A link between per capita fat intake and breast cancer risk is also supported by a number of inter- national studies. although case-control studies do not. in general. support this link. Animal studies have shown that high levels of fat in the diet enhance the development of both spon- taneous and chemically induced mammary cancers in mice. This was true even if the fat was fed after tumor initiation. lending support to the hypothesis that dietary fat acts as a cancer promoter. 45 46 Correlations of international incidence and death rates with diet components indicate that colon can- cer and. to a lesser extent. rectal cancer. are associated with total dietary fat intake. Some case- control studies have also suggested a relationship between dietary fat intake and risk of colon and rectal cancers (Palmer and Bakshi. 1983). but the evidence is not clear—cut. Varying levels of fiber intake may help to explain some of the conflicting data on fat intake and colon and rectal cancers. because fiber is thought to pro- tect against these cancers. Thus. fat intake—par- ticularly milk fat—was found in one study to be about the same for individuals in rural Finland and in Copenhagen. Denmark. but the Finns. who eat large amounts of high-fiber. unrefined rye bread. had a much lower incidence of colon cancer (MacLennan et al. I978). An association between the incidence of prostate cancer and high dietary fat intake in humans has been seen chiefly in international surveys. Both prostate cancer incidence and fat intake are higher among Japanese living in Hawaii than among Japa- nese living in Japan (Waterhouse et al. 1976). And prostate cancer incidence has been linked with the consumption of animal fat and protein among sev— eral ethnic groups in Hawaii (Kolonel et al. 198]). Most population studies of dietary fat intake and cancer incidence have not been able to differenti- ate the effects of saturated and unsaturated fats. but consumption ofthe two often go hand in hand. Total fat intake accounts for about 40 percent of total calories in the US. diet. compared with only 20 percent in the Japanese diet. Meat. eggs. dairy foods. and oils used in cooking and in salads are the chief sources of fat in the American diet. Fiber Dietary fiber appears to protect against some forms of cancer. particularly colon cancer. Which types of fiber. and how they may work. are not clear. Fiber is found in fruits, vegetables. nuts. legumes (peas and beans). brown rice. and whole— grain breads and cereals. Some studies have shown an inverse association between the consumption of vegetables and the occurrence of colon cancers (Graham. 1972: I978). But it has been difficult to single out the role of fiber in many of the studies that measure total fruit and vegetable intake. lndi~ viduals who eat diets high in fruits. vegetables. legumes. and whole grains usually eat less fat and protein and tend not to be overweight. And a diet high in fruits and vegetables is likely to be high in carotenoids and vitamin C. If some fiber component does protect against colorectal cancer. it may do so by speeding the transit time of fecal matter through the bowel. thus decreasing the time that food carcinogens stay in contact with bowel walls: by increasing the bulk of the stool and thus decreasing the concentration of carcinogens; or by changing the different bacterial species in the bowel. some of which may destroy carcinogen metabolites (Doll and Peto. I981). It has also been postulated that dietary fiber intake decreases the production of mutagens in the stool (Ehrich. 1979). Cured, pickled, and smoked foods; molds and fungi Studies in different parts of the world show an ap- parent link between frequent consumption of pick- led. cured. and smoked foods and an increased risk of stomach cancer. This association could be due either to the curing agents or to molds and fungi toxins that may form in these foods if they are not refrigerated. The decrease in stomach can- cer incidence in the United States since the 1930s has been attributed to the wide use of refrigeration. although there is no proof. Vitamin E and Selenium Both vitamin E and an enzyme that depends on se- lenium for its activity act as antioxidants. able to block damage to cellular DNA from some car- cinogens. These substances can also keep cultured cells treated with chemicals from becoming can- cerous. There is no evidence that “megadoses” of either of these substances help to protect humans against cancer. Selenium is known to be toxic in high doses. and vitamin E. which is fat—soluble. is potentially toxic. 47 Risk Factors Obesity Obesity has been associated with increased cancer death rates in animals in a number of studies over the past 45 years (Doll and Peto. I981). It is also suspected that obesity is associated with increased death rates for some cancers in humans. particu— larly those ofthe prostate. pancreas. breast. and ovary. One large study. ofthree-quarters of a mil- lion men and women observed over a 13—year period (Lew and Garfinkel. 1979). has lent some support to these suspicions. Drugs 48 The development of "miracle“ drugs that effective— ly treat a variety of illnesses has been one of medi- cine‘s major achievements. Unfortunately, when chemically altering or arresting the course of one disease. these drugs can contribute to the develop- ment of other diseases. including cancer. Estro- genic hormones. immunosuppressive agents. and, ironically. anticancer drugs designed to kill tumor cells are the classes of drugs most often linked to human cancer. Drugs are believed to account for fewer than 2 percent of all cancers. When deciding to use drugs. the informed patient and physician must carefully weigh the benefits of a medication against its possi- ble risks. Hormones Estrogens are hormones. produced mainly by the ovaries. that help regulate menstruation and preg- nancy. They are also responsible for the develop- ment of feminine body features. The ovaries pro- duce several different estrogens—estradiol, estriol. and estrone—but because they have similar func— tions and chemical structures. they are collectively referred to as estrogens. Many middle—aged and older women take syn- thetic “replacement“ estrogens to relieve symp- toms of menopause that may develop when their ovaries decrease production of these hormones. Such use has been linked to cancers ofthe endo- metrium. the lining of the uterus. One study reported that women who take replacement estrogens for more than 7 years increase their risk of endometrial cancer 14 times (Zeil and Finkle. 1975). It appears. though. that women who stop taking replacement estrogens decrease their risk of this cancer to the level of a nonuser after 2 years (Stolley and Hibberd. 1982). It is not clear if replacement estrogens increase the risk of breast cancer in menopausal women. Several studies have shown a slightly increased risk (Hoover et al. 1976; Brinton. I981: Ross. I980: Thomas. 1982: Hoover. 1981; Jick and Walker. 1980: Hulka and Chambless. 1982) especially in long-term users. The groups these studies looked at differed, and other investigators have failed to show an association (Kelsey et al. 198]). Thus. results are not conclusive. Synthetic estrogens are the chief ingredient of birth control pills. The most effective and most widely used birth control pills are “combination" pills that contain both estrogen and a second ovarian hormone. progestin. Pills that contain only progestin. called “minipills.” are also available. Combination pills may actually decrease a wom— an's risk of developing some cancers. Women who take these pills are only half as likely as nonusers to develop cancer of the ovary or endometrium: their use is estimated to prevent 1,700 cases of ovarian cancer and 2,000 cases of endometrial can- cer each year (CDC. 1983). A different form of birth control pills. known as “sequential" pills, were available at one time. They were taken off the market by the US. Food and Drug Administration in the late l970s because stud- ies linked them to an increased risk of endometrial cancer. Sequential pills provided separate estrogen and progestin in sequence: during the first weeks of her menstrual period. a woman took the estro- gen pills. and in the last week took the progestin. One study showed a sevenfold increase in risk of endometrial cancer among women who took one brand of sequential pills, but failed to find an in- creased risk among users of other brand (Weiss and Sayvetz. I980). The risk was associated with the brand that contained the most potent form of estrogen and the weakest progestin. 49 50 The link between birth control pills and breast cancer is less clear. Most ofthe evidence suggests that women who use oral contraceptives do not have an increased risk of breast cancer despite duration of pill use and other variables. In some studies. though. specific groups of women who use these pills have been adversely affected: those with a family history of breast cancer or those with be- nign breast disease. A recent study suggested a fourfold increase in breast cancer incidence among women under age 36 who were long-term users of combination pills containing a “high" dose of pro- gestin (Pike et al. 1983). Other recent studies have linked oral contraceptives with an elevated risk of cervical cancer. even after sexual activity and other risk factors for this cancer were taken into account. DES (diethylstilbestrol). an estrogen-like com- pound. has been widely publicized as a cancer- causing drug. During the [940$ and l950s. doctors prescribed DES to pregnant women because they believed the drug helped prevent miscarriages. Al- though studies in the 1930s had shown that DES caused cancer in laboratory animals. the problem received little attention until the 1970s. when a rare vaginal and cervical cancer was found in a number of young women living in areas where DES had been widely prescribed. Histories showed that the patients had been exposed to DES in II[(‘I'() when their mothers took the drug (Herbst et al. 197]). It is now thought that between four and six million Americans—mothers. daughters. and sons—may have had pregnancy-related exposure to DES. The DES findings also provided the first evidence that a carcinogen could travel across the placental bar- rier from the mother to the fetus. There have been few studies of the effects of DES on women who took it. and those studies pro- duced conflicting results. One study showed some increased risk of cancer of the breast and repro- ductive organs in these women (Bibbo et al. I978); another showed no such link (Brian et al. I980). DES has not been associated with cancers of the reproductive tract in sons of women who took it. but anatomic abnormalities of the sperm and re- productive tract have been reported in these men. Studies to see if these abnormalities affect their fertility have been delayed because. as a group. these DES sons seem to have postponed starting families (DES Task Force. 1978). Androgens. a class of male hormones. are struc- turally related to estrogens. Synthetic androgens are used to treat various cancers, some genetic dis- eases. and some blood and endocrine disorders. Some athletes use androgens to increase body mass and athletic performance. Studies have shown an increased incidence of liver cancer among children who received andro— gens for the treatment of some types of anemias (Johnson et al. I972). It is not known. though. if the liver cancer these children developed resulted from the androgens or the disease. Anticancer drugs In the last decade. anticancer drugs have pro- longed the lives of many thousands of cancer patients. A number of these patients have devel- oped second cancers. Alkylating agents are a class of drugs used to treat a variety of cancers. They work by inserting foreign molecules into the genetic material of divid- ing cancer cells. These foreign molecules act like wrenches thrown into the machinery of the cell. preventing growth and division. Alkylating agents may also cause mutations in the cell's genetic material. similar to the mutations caused by ionizing radiation. and lead to cancer. Several studies have shown that Hodgkin‘s disease patients treated with alkylating agents have an in- creased incidence of acute myelocytic leukemia (AML). Not a single case of AML was found in more than 3.000 Hodgkin's disease patients studied be— fore 1962 (Moertel and Hagedorn. 1957 and Berg. 1967). But a study done between l96l and I973 showed that Hodgkin's disease patients were devel— oping AML at a rate of 156 cases per 100.000 popu- lation (Weiden et al. 1973). The greatest incidence of AML in this 12-year study occurred from 1970 to 1973. a few years after the MOPP regimen (ni- trogen mustard. vincristine [Oncovin]. prednisone. and procarbazine) was introduced in I967. Both nitrogen mustard and procarbazine are alkylating 51 agents. The risk of leukemias after treatment of Hodgkin's disease has continued to increase. espe- cially among patients who receive both MOPP and radiation therapy. Alkylating drugs used to treat other cancers have also been linked to cancer. Epidemiologic studies showed a hundredfold increase in AML among multiple myeloma patients after the drug melphalan was introduced (Adamson and Sieber. 1977). A recent survey of women treated for ovarian cancer showed that melphalan and chlor- ambucil increase the leukemia risk in steps that correlate with dose (Greene et al. 1982). The risk of acute leukemia has also been shown to rise after use of alkylating agents to treat lung cancer (Stott et al. I977). the blood disorder known as polycythemia vera (Berk et al. I981). and non- Hodgkin‘s lymphoma (Greene et al. l983). The alkylating agent methyl-CCNU (semustine). used to treat cancers of the stomach. colon. and rectum. was recently shown to increase a patient's risk of developing leukemia by sixteenfold (Boice et al. I983). Most surveys of chemotherapy-related cancers have focused on leukemia because it devel— ops within 4 to 5 years after exposure. The risk of other cancers has not been evaluated in a systemat- ic way because most exposed patients have not lived long enough to develop other cancers. lmmunosuppressive drugs Drugs that suppress the immune system are given to organ transplant patients to help them accept a foreign organ. Azathioprine and adrenal cor- ticosteroid hormones were widely used in the past for kidney transplant patients. and cyclosporin A has been used recently. Non-Hodgkin‘s lymphoma. the most common type of cancer in these patients. occurs 32 times more often among them than in the general population (Hoover and Fraumeni. 1973). The lymphoma often arises rapidly—within a year or two—after the transplant operation. and often develops in the brain. an unusual site for this type of cancer (Hoover. 1977). Some other cancers—— skin cancers. Kaposi‘s sarcoma, and lung cancer— also occur at a high rate in transplant recipients. Risk Factors Other Drugs Radioactive drugs contain a molecule "tagged" with a radioactive isotope so the isotope can be counted or imaged in diagnostic tests. Radioactive drugs can concentrate in body tissues and. depend- ing on their strength and half—life. may injure those tissues. Radioactive drugs are also used to treat bone tuberculosis. thyroid cancer. and the blood disorder polycythemia vera. Some of these radio- active drugs have been shown to cause various can- cers. including osteogenic sarcoma. a type of bone cancer: leukemias: and a rare form of liver cancer (Hoover and Fraumeni. 1981). In 1964. chlornaphazine. a drug used to treat polycythemia vera and Hodgkin's lymphoma. was withdrawn from the market because it was found to cause bladder cancer. Chlornaphazine is chem— ically related to beta-naphthylamine, a chemical earlier associated with bladder cancer among workers in the dye industry. Other drugs have also been found to increase the rate of human cancers. Pain-killing drugs that con— tain phenacetin have been linked to kidney can- cers, and methoxypsoralens. used with ultraviolet- A radiation in the PUVA regimen for psoriasis. have been linked with skin cancer. Familial Factors Most cancers are caused by a variable mix of hereditary and environmental factors. Some rare cancers are inherited. and usually appear at an ear- ly age. A number of rare hereditary disorders may predispose a person to cancer but the added action of one or more environmental factors is often needed for the cancer to develop. Other individuals seem to be resistant to some cancers. Subgroups of some of the common cancers have a genetic com- ponent but may also require an environmental trig- ger. Some nonhereditary cancers seem to run in families. but this may reflect chance or a common environmental exposure. How a person reacts to his environment is also a part of the equation. As an example of the interactions of genetic and environmental factors. consider the Australian 53 54 woman who develops nonmelanoma skin cancer. She inherited her fair skin; that is the genetic com- ponent. She lives near the equator. where sunlight is intense and prolonged; that is an environmental factor. She fails to avoid the sun at noontime or to protect her skin with clothing and sun screens. and that is yet another factor: how the individual inter- acts with her environment. This mix of factors can apply to the origins of many cancers. and suggests various possible preventive measures. Retinoblastoma. an eye cancer. is one ofthe very rare inherited cancers. About 40 percent of all reti- noblastoma cases are inherited. and occur in both eyes during early childhood. The non—heritable cases usually occur in one eye only. and at a later age. Predispositions for certain rare cancers also are inherited but seem to require an additional en— vironmental event (Miller. in press). These include the hereditary form of Wilms' tumor. a childhood kidney cancer; xeroderma pigmentosum. which makes individuals extremely sensitive to sunlight and to nonmelanoma skin cancer; dysplastic nevus syndrome. a precursor of melanoma; and ataxia- telangiectasia. an inherited immune deficiency syn- drome that predisposes individuals to lymphoma. leukemia. and stomach cancers. Certain “cancer families" have also been identi- fied and followed. The pattern oftheir cancer occurrences indicates that the cancers were inher- ited. though the family members may develop dif— ferent forms of cancer (Li and Fraumeni. I982). Subgroups of breast cancer. colon cancer. and brain cancer also appear to be heritable. And with some of the other common cancers. such as stom- ach cancer and lung cancer. a familial component is strongly suspected. though it could reflect ex- posure to a common cause. Thus. the risks of developing breast cancer or colon cancer may be twenty- to thirtyfold greater than normal in per- sons with familial risks. In one family with kidney cancer. followed for four generations. the youngest members are being closely watched because their chances of developing it are so great. Some ethnic groups seem to possess traits that protect them against some cancers. Thus. chronic lymphocytic leukemia is extremely rare among Risk Factors Orientals. and Ewing’s sarcoma is very rare among blacks. Genetic “markers" help to identify family mem- bers who are at high risk of developing certain heritable cancers so that they can be watched for signs of early disease. A “marker" for retinoblasto— ma. for example. is a missing chromosome seg- ment. In instances where a predisposition is inher- ited. care can be taken to prevent exposure to the environmental event or “insult" that causes cancer. Removal of precursor moles for melanoma. or of polyps for colon cancer. are ways to prevent devel— opment of disease. even though the predisposition for it has been inherited. Meanwhile. on another front. basic researchers are trying to learn exactly what happens at the level of the gene. Recent oncogene research has led to the hypothesis that normally “quiet" genes found in all vertebrate cells may be triggered—by viral infection. by environmental events or by some other mechanisms—at some step in the can- cer process. Ionizing Radiation Ionizing radiation is a known cause of cancer. and of other adverse effects as well. It is one of the most extensively studied human carcinogens and may account for about 3 percent of all cancers. Ionizing radiation is able to remove electrons from atoms and to change the molecular structures of cells. It is these cellular changes that may cause cancer to develop. The genetic DNA in the cell nucleus is thought to be the critical target for radiation-induced damage. Some radiation comes from natural sources. like that from cosmic rays and from radioactive sub- stances in the earth‘s crust. Each of us is exposed to this “background“ radiation at a rate of about 0.1 to 0.2 rad per year. (The rad is a unit of mea- surement for the amount of radiation energy absorbed by body tissues.) Very high doses of radiation (hundreds of rads) received all at once may be fatal. but if spread out over a period of time, a high dose of radiation may be less damaging to healthy tissues. A single. 500- 55 56 rad dose of whole-body irradiation, for example. would cause about half the people exposed to it to die within 30 days. But patients who receive daily radiation therapy treatments of 200 rads a day di- rected to a small area of the body can be exposed to thousands of rads over a period of weeks. It is difficult to measure the effects of low-level radia- tion—of less than 25 rads, say—from common sources like medical X-rays. Some of what we now know about the effects of radiation exposure was learned by studying pa- tients who received medical treatments with radia— tion in the past. Some of the first quantitative evidence that radiation causes cancer came from studies of patients who received radiation therapy before 1954 for a spinal disorder (Court Brown and Doll, 1965; Smith and D01], 1982). Researchers noticed that these patients had more leukemia and cancers of the stomach, pancreas, lung, and other highly irradiated organs than would be expected in a healthy population. Between 1935 and 1954, fluoroscopy, an X-ray procedure, was used to monitor treatment of pa- tients with tuberculosis; the women who were thus treated could receive an average dose of 150 rads to the breast. Ten to 15 years later these women had a high incidence of breast cancer (Boice and Monson, 1977). There is a high risk of thyroid cancer many years after childhood exposure to radiation for therapy of noncancerous conditions of the head and neck. Ex- posure during childhood can be particularly dam- aging because rapidly growing cells may be more sensitive than slower growing cells irradiated later in life. Before 1950, individuals with enlarged thy- mus glands were treated with intense radiation. An elevated risk of thyroid cancer and leukemia has since been found in this population. Females were at greater risk than males, and the risk among Jews was especially high (Hempelmann et al. 1975). Those who think they may have received radiation treatments for thymus conditions or other child- hood problems, such as scalp ringworm or en- larged tonsils, should tell their physicians. The radiologists who first used X-rays exten- sively before 1920, when the potential dangers of radiation were not recognized, had high rates of leukemia, a blood cancer. Today. precautions are taken to reduce the exposure to patients, radiol- ogists, and X—ray technicians. A classic study of women who painted radium dials before 1930 showed high rates of bone sar- comas and head cancers. They had swallowed large quantities of radioactive radium by licking their paint brushes to make fine tips; it has been esti- mated that the average dose reaching their bone tissues was 1700 rads (BEIR, I972). Most of our information about the effects of radi- ation comes from studies of atomic bomb survivors in Japan. among whom have been found increased rates of acute leukemia and cancers of the breast. thyroid. lung. stomach, and other organs (Boice and Land, 1982). Female survivors who received a single dose of radiation from the blast were found to be at the same risk for breast cancer as the women with tuberculosis who had repeated fluor- oscopy exposures over a 3- to 5-year period (Boice et al. 1979). This suggests that in the case of breast cancer—but not necessarily other cancers—repeat- ed small doses over the years may be as hazardous as a single. large dose. Today, many women at risk of developing breast cancer are periodically examined with mammogra- phy. or low-dose breast X-rays. But for high-risk women. particularly over age 50. the benefits of detecting cancer early outweigh the small risk of developing cancer from repeated mammograms. Exposure to low levels of radiation before birth is associated with the development of cancer dur— ing childhood. especially leukemia (Bithell and Stewart, 1975). but not all researchers are con— vinced that prenatal irradiation is the cause of childhood cancer. Individuals exposed prenatally during the atomic bomb blasts in Japan do not have higher cancer rates. Now. though, ultrasound is used during pregnancy instead of X-rays when possible. There are other environmental and occupational exposures to radiation. Radioactive fallout. for ex- ample. is produced during nuclear weapons tests when airborne radioactive particles settle to the ground. One study showed that persons acciden- tally exposed to very high levels of fallout had an increased risk of thyroid cancer. and that females 57 Risk Factors were especially susceptible (Boice and Land. [982). Uranium miners inhale radioactive radon gas produced underground by the natural decay of ura- nium. and have high rates of respiratory cancer (Archer el al. 1976). There may also be an interac- tion between inhaled radioactive gas and smoking that increases the risk (Wittenmore and McMillan. 1983). Radon is becoming a public health concern in some locations because of its presence in ground water and building materials. In general. the breast. thyroid. and bone marrow are most sensitive to the effects of ionizing radia- tion. There may be a minimum lag time after ex- posure of about 2 years before leukemia develops. and 10 to 15 years before other cancers develop. Reducing exposure to unnecessary medical X- rays is one of the best ways to reduce exposure to ionizing radiation. In many instances, though. the benefits outweigh the risks. as in mammography for some women, as a tool for diagnosis of various diseases or injuries. and as a way to treat some cancers. Occupation 58 In I775. London surgeon Percivall Pott published a report on a rare cancer of the scrotum that he found was common among chimney sweeps. a dis- ease known at the time as “soot-wart" (Shimkin. I977). A century later. two other scientists noted similar cancers among gas plant workers in Ger- many and oil shale workers in Scotland. Yet an- other 40 years later. certain constituents of tar. soot. and oils. known as polycyclic aromatic hydrocarbons. were found to cause cancer in laboratory animals. This briefdetective story has become a model for many later investigations of workplace carcino- gens: observation of unusual cancers. or a high in— cidence of more common cancers. among groups of workers; a search for a responsible agent: and finally. demonstration that the agent can cause can— cer in laboratory animals. A postscript to the chimney sweep story deals with the ultimate goal of workplace cancer studies: prevention. Spurred by Pott's report. the Danish chimney sweeps guild in 1778 urged its members to take daily baths (Clemmesen. 195]). A report in the l892 British Medical Journal. “Why Foreign Sweeps Do Not Suffer From Scrotal Cancer." noted that the sweeps of northern Europe seemed to benefit from this hygiene measure. but English sweeps. who apparently ignored such implications. continued to develop cancer. Industrial workers are. in a sense. flagmen for our society. In this age of chemicals. metals. plas- tics. and fibers. we all run the risk ofexposures to industrial carcinogens in our air. water. food. and homes. But the exposures of the industrial worker may be intense and prolonged. If a substance used in the workplace is a carcinogen. the cancers it can cause will most likely be seen first in workers. Since l97l. the International Agency for Re— search on Cancer ([ARC). an agency of the World Health Organization headquartered in Lyons. France. has been publishing critical reviews of data on the carcinogenicity of chemicals to which hu- mans are exposed. and [ARC working groups of scientists have been evaluating these data in terms of human risk. A summary of these reviCWs was published in 1982 (lARC. Supplement 4). The [ARC scientists assessed the data from epi- demiologic studies of humans. from studies in ex- perimental animals. and from short-term tests or assays. Based on these data. they assigned individ— ual chemicals. chemical groups. industrial pro- cesses and occupational exposures to one ofthree categories of risk. (See tables on next two pages.) When there was enough evidence from epi— demiologic studies to support a causal association with cancer. the chemical. chemical group. pro- cess. or exposure was assigned to Group 1. The scientists placed the chemicals. chemical groups. processes. or exposures that they considered prob- ably carcinogenic to humans in Group 2. Those with data to support the highest carcinogenic risk to humans appear in Group 2A: those with lower risk are in Group 28. The carcinogenic agents included in the tables do not fall into any particular class of substances: they may be metals. solvents. dyes. organic and in- organic dusts. pesticides or herbicides. Space here 59 TABLE I Group 1: 60 Industrial processes and occupational exposures causally associated with cancer in humans: Auramine manufacture Boot and shoe manufacture and repair (certain occupations) Furniture manufacture Isopropyl alcohol manufacture (strong~acid process) Nickel refining Rubber industry (certain occupations) Underground hematite mining (with exposure to radon) Chemicals and groups of chemicals causally associated with cancer in humans: 4—Aminobiphenyl Arsenic and arsenic compounds Asbestos Benzene Benzidine N,N-Bis(2—ch|oroethy|)-2-naphthylamine(Chlornaphazine) Bis(chloromethy|)ether and technical—grade chloromethyl methyl ether Chromium and certain chromium compounds 2»Naphthy|amine Soots. tars and oils Vinyl chloride TABLE II Group 2A: Acrylonitrile Benzo(a)pyrene Beryllium and beryllium compounds Diethyl sulphate Dimethyl sulphate Manufacture of magenta Nickel and certain nickel compounds ortho-Toluidine Group 28: Amitrole Auramine (technical grade) Benzotrichloride Cadmium and cadmium compounds Carbon tetrachloride Chlorophenols DDT 3,3’-Dich|orobenzidine 3,3’-Dimethoxybenzidine (ortho-Dianisidine) Dimethylcarbamoyl chloride 1,4-Dioxane Direct Black 38 (technical grade) Direct Blue 6 (technical grade) Direct Brown 95 (technical grade) Epichlorohydrin Ethylene dibromide Ethylene oxide Ethylene thiourea Formaldehyde (gas) Hydrazine Phenoxyacetic acid herbicides Polychlorinated biphenyls Tetrachlorodibenzo—para-dioxin (TCDD) 2.4.6—Trichlorophenol 61 is too short for a detailed discussion ofeach of the agents in Tables I and ll. and such information can be found in the [ARC series. But to illustrate briefly how exposures occur. how risk is ascertained. and how exposures can be minimized or prevented. four types of workplace carcinogen. each from Group I. are discussed here in some detail: a mineral fiber (asbestos). a chem— ical (benzene). a metal (chromium). and an indus— trial process (furniture manufacture). Asbestos Derived from the Greek word meaning “incom- bustible." asbestos is the generic name fora group of minerals composed of silicate fibers that are heat-stable. non—conductive and can be woven. As- bestos occurs widely in mineral formations found throughout the world and has been used since the late l8()()s. ln I976. more than 5.500 tons of as— bestos were produced (IARC. Vol I4. 1977). the bulk of it by Canada and the Soviet Union. About two-thirds of the asbestos produced is used in the construction industry for cement sheets and pipes. insulating materials. and floor and ceiling tiles. It is also used for friction applications. like clutch and brake l‘acings for cars and machinery. It is used widely in the shipbuilding industry. chiefly for in- sulating and lire-proofing. In the past. it was often sprayed on for insulation. fire-proofing. and deco- rative and acoustic uses. but its use in many of these settings is now decreasing. The presence of asbestos in various materials does not necessarily pose a health risk. The health risk arises when the asbestos fibers are set free during mining. drilling. sawing. or spraying. or when materials with asbestos in them start to decompose. Reports linking asbestos with lung cancer and asbestosis. a lung disease. have been appearing since I935. The earlier studies reported these dis— eases mostly among asbestos workers and miners. but as the uses of asbestos multiplied. particularly in shipbuilding during World War II. the numbers of workers who suffered the injurious effects of as- bestos exposure continued to grow. Mesothelioma. a rare cancer that affects the lining ofthe chest and abdominal cavity. has been linked with asbestos exposure since the early 1960s. Inhaled asbestos fibers produce these cancers in a number of differ— ent laboratory animals. Lung cancer and mesothelioma both have laten— cy periods of up to 30 years or more: this means it may take that long for cancer to develop after ex- posure. But the duration ofthe exposure needed to cause cancer may be very brief. In some cases. in- dividuals who were exposed to asbestos for only a few months have developed cancer. Men who worked in shipyards for only a few years during World War II. for instance. have been found to be at high risk of lung cancer (Blot et al. l982). Some cases of lung cancer and mesothelioma have also been found to occur among "bystanders": the fam- ilies of workers who carried asbestos fibers home on their clothing: workers who were nearby when asbestos materials were sawed. drilled. or sprayed: individuals who lived near asbestos mines or facto- ries where asbestos was fabricated. Cigarette smoking and asbestos exposure have a synergistic effect: smoking plus asbestos exposure creates far more risk than either one alone. Some other cancers have also been linked to asbestos ex- posure. including cancers of the esophagus. stom- ach. colon. larynx. and kidney. Asbestos exposure is now regulated by the ()c- cupational Safety and Health Administration (OSHA). which sets standards for the number of fibers that may be present in the workplace air. It also requires personal protective equipment for workers. The National Institute for Occupational Safety and Health (Nl()SH). the research arm of OSHA. recommends that all non-essential uses of asbestos be eliminated. Benzene A clear. colorless liquid. benzene was first isolated by Faraday in 1825 from a liquid condensed by compressing oil gas (lARC. Vol 7. I974). It is now produced from petroleum. Up until World War ll. benzene was used chiefly in this country as a gas- oline additive. Now it is also used widely as a sol- vent in the chemical and drug industries. and as an intermediate reactant in the synthesis of resins. 63 64 adhesives. and plastics. Case reports linking leukemia with occupational exposure to benzene have appeared in the liter— ature since the late l920s: such reports have come from a number of countries. The major route of ab- sorption of benzene is through the lungs: most of the affected workers had breathed its fumes. Among those exposed who later developed leuke- mia were workers who made artifical leather. or made shoes using rubber cements that contained benzene. Case—control and cohort studies (lARC. Vol 7. 1974) have supported this association. Benzene causes chromosomal abnormalites. but not mutations. in some short-term assays (lARC. I982) and has recently been found to be carcino- genic in assays with laboratory animals (NTP. I984). Some two to three million workers may be ex~ posed to benzene in petrochemical. rubber. and coke plants. Shoemakers. furniture finishers. and gas station attendants are also exposed to it. OSHA regulates workplace exposures. Chromium and chromium compounds The element chromium is the 20th most abundant element and occurs. mostly in chromite ores. throughout the world. The Republic of South Af- rica and the Soviet Union are now the two major producers. Chromium and its compounds are used widely in industry because of chrome's hardness. resistance to corrosion. high melting point. and wide availability. Its name is derived from the Greek word “chroma.“ for color. because of the reds. oranges. yellows. and blues of its salts when it is combined with other minerals. Most chromium compounds are used in the met— allurgical industry. particularly to make stainless steel and alloys. Chromium compounds are also used in the manufacture ofbricks. glass. ceramics and certain iron-containing metals. The colored chromate salts of various metals. like zinc and po- tassium. are used in the pigment. paint. tanning and dyeing industries (IARC. Vol 23. 1980). An increased incidence of lung cancer has been seen among workers in the chromate-producing industry. There are also indications of high risk among chromium platers and chromium alloy workers. Furniture manufacture For many centuries. furniture and cabinet-making were performed by craftsmen and artisans working at home or in small shops. With the advent ofthe Industrial Revolution and the development of cir- cular saws and planers. furniture and cabinet-mak- ing became an industry. Advances in mechaniza- tion were most rapid after World War I. With the development of bandsaws. routers. lathes. and sanders. furniture-making became an assembly-line process, a process that created more furniture and more hazards: high noise levels. accidents. and dust pollution. It was not until the l960s that the cancer-causing effects of furniture dust began to be recognized. An increased incidence of cancers of the nose and nasal cavities was first reported among fur- niture makers in England. Later reports came from Italy. Canada. the Netherlands. Denmark. France. Germany. Sweden. and other parts of England (IARC. Vol 25. l98l). A study in this country (Brinton et al. 1984) showed high nasal cancer risks among furniture workers. particularly in North Carolina. There have also been reports ofhigh inci- dence of cancers of the larynx. lung. and gastroin- testinal tract among furniture workers in the United States. England. and Germany. although these associations are not clear. From the epidemiologic data available. [ARC has concluded that nasal adenocarcinomas have been caused by employment in the furniture-mak- ing industry. and that the excess risk of these can— cers occurs mainly among those workers exposed to dust from certain hardwoods. Suction devices. particularly near saws and sanders. can markedly reduce the amount ofair- borne dust at the workplace (Hounam and Wil— liams. I974) and masks worn by workers can further reduce the amount of dust inhaled. The remaining chemicals or processes listed in Tables I and ll each fall into the four categories discussed above. Ways to reduce workplace ex- posure to them. including labeling. venting. and in- 65 Risk Factors dividual protective measures, are regulated by a number of Federal agencies. NCl studies are at- tempting to gain further understanding of work- place carcinogens, and are looking at ways to counter the cancer risks of workers who have already been exposed to workplace carcinogens. Also, some industries are testing substances in short-term cell-culture assays before introducing them to the workplace. Solar Radiation 66 Solar radiation is the chief cause of nonmelanoma skin cancer, responsible for about 90 percent of cases. It has also been linked with skin melanoma, but that relationship is more complex. Though nonmelanoma skin cancers are now considered to be 98 percent curable, they still accounted for as many deaths in the United States during the 19505 and 1960s as did melanomas, which are far rarer but far more lethal (Mason et al, 1975). More than 400,000 new cases of non— melanoma skin cancer are thought to occur in the United States each year, and this number is rising. Nonmelanoma skin cancer is the most common form of cancer among Caucasians. The relationship between sun exposure and non- melanoma skin cancer has been clarified greatly in the past decade. Observers had noted in the late 1800s that sailors exposed to the sun developed “Seamanshaut,” or “sailor’s skin," and in the early 1900s an excess risk of skin cancer was observed among farmers. The greater risk for Caucasians ex- posed to sun had also been observed (Hyde, I906). By I928, scientists were able to demonstrate the cancer-causing effects of ultraviolet radiation on the skin of lab animals, using both sunlight and artificial light sources (Findlay, 1928). These car- cinogenic effects were produced by ultraviolet-B (UV-B) radiation in the 290 to 320 namometer range, the same range that produces burning in human skin. Though latitude, or distance from the equator. generally determines the amount of UV—B radiation in a given location, altitude and sky cover are also determining factors. Atlanta, Georgia, and El Paso, Texas, for example, are in the same general latitude (32 to 33° N). But El Paso, which is higher and drier, has an annual UV—B count 38 percent higher than Atlanta. Time of day and time of year also affect the amount of UV—B radiation in any location. The greatest amount, of course, occurs during the sum- mer months, and a third of the day’s total amount occurs between the hours of 11 a.m. and 1 pm. (or 12 noon and 2 pm. DST). In 1973, special meters designed to measure UV-B radiation were set up in a number of US. cities by Temple University in collaboration with the National Cancer Institute and other govern- ment agencies. Data from these meters permitted precise correlations, for the first time, with NCI data on skin cancer incidence in four of these cities, thus affording observations on human popu- lations (Scotto et al, 1976). The most striking association was the inverse re- lationship between latitude and nonmelanoma skin cancer incidence: the lower the latitude (the equa— tor is at zero), the higher the incidence. The data also indicated that nonmelanoma skin cancer is re- lated to annual, cumulative UV—B exposure, while skin melanoma may be related to brief exposure to high-intensity UV radiation. Other findings from the UV—B stations and NCl incidence data are that in some parts of the South, the incidence of skin cancer exceeds that of all other cancers combined, and in parts of the North it accounts for about 40 percent of all cancers. Another factor that affects UV-B exposure is the amount of ozone (O3) in the atmosphere (Scotto et al, I982). Ozone gases absorb most of the UV light in the upper stratosphere and let only small amounts reach the earth. There has been concern in the past decade that the exhaust gases of super- sonic aircraft, some fluorocarbons (particularly those used in spray propellants), nuclear weapons, and nitrogen fertilizers could deplete this ozone layer. A Federal task force warned in 1975 that if 1972 levels of fluorocarbon release were continued, the ozone concentration would be reduced by up to 7 percent within several decades. It also ob- served that there would be about a 2 percent in- crease in UV—B radiation near the equator for each 67 Risk Factors 1 percent reduction in stratospheric ozone con- centration. These concerns are now being studied by a number of Federal agencies. Solar radiation ranks high among the “lifestyle" factors associated with cancer. Most individuals have some choice in the amount of sunlight ex- posure they get. and too much sun is the chief cause of nonmelanoma skin cancer. Tobacco 68 Cigarette smoking is the single major cause of can- cer mortality in the United States. Cigarette smok— ers have total. overall cancer death rates two times greater than nonsmokers; heavy smokers (over a pack a day) have a three- to four-times greater risk. If every smoker in this country were to stop. the overall cancer death rate in the United States would decline significantly. There is no single action an individual can take to reduce the risk of cancer more effectively than to quit smoking. par— ticularly cigarettes (U.S. Surgeon General’s Report, 1982). The American Cancer Society esti- mated that tobacco use contributes to more than 350,000 premature deaths each year (ACS. I983). Cigarette smoking increases the risk of cancer of the lung more than any other site. But cigarette smoking—as well as pipe and cigar smoking—also multiplies the risk of cancers of the lip. mouth. tongue, and pharynx. depending on the type and amount of smoking (Fraumeni, I975). Cigarette smoking also increases the risk of cancers of the larynx and esophagus. Heavy drinking of alcoholic beverages, combined with smoking. increases the risk of cancers of the mouth and throat. larynx, and esophagus even further. Cigarette smoking is also linked with increased incidence of cancers of the bladder. pancreas, and kidney, though to a lesser degree than with cancers of the respiratory and upper digestive tracts. In ad- dition, some epidemiologic evidence points to an association between cigarette smoking and stom- ach cancer, and there is conflicting evidence about the role of cigarette smoking and cancer of the uterine cervix (U.S. Surgeon General’s Report. 1982). 1‘ Cancer death ratios among men who smoked cigarettes compared with a risk of 1.0 . for men who never smoked regularly (American Cancer ' Society Survey, 1966) Age 45—64 Age 65—79 Total Cancer 2.14 1.76 Lung 7.84 11.59 Mouth, pharynx 9.90 2.93 Larynx 6.09 8.99 Esophagus 4.17 1.74 Bladder 2.00 2.96 Kidney 1.42 1.57 Prostate 1.04 1.01 Pancreas 2.69 2,17 There are distinct dose-response relationships between cigarette smoking and cancers of the lung. mouth and throat. esophagus, bladder. and larynx; the risk of developing these cancers increases with the number of cigarettes smoked each day. Those who smoke filtered. low-tar cigarettes have a lower lung cancer risk than those who smoke nonfiltered, high-tar cigarettes. The cancer risk for such smok- ers is still far higher than for nonsmokers, however (Wynder. 1982). Pipe and cigar smoking. tobacco chewing. and snuff dipping also increase the risk of certain can- cers. Both pipe and cigar smoking account for some lung cancer risk (though far less than ciga- rette smoking). and both increase risk for cancers of the mouth. esophagus. and larynx (U.S. Sur- geon General’s Report. 1982). Snuff dipping, common in parts of the South and on the increase among teenagers. increases the in- cidence of cancers of the mouth and throat. (Snuff dipping is the practice of holding a cud of finely ground tobacco in the cheek.) The risk of cancers of the cheek and gum has been shown to be in- creased nearly 50—fold in long-term snuff users (Winn et al. 1981). The number of cancers of the mouth and esophagus associated with tobacco chewing and snuff dipping is relatively small in this country. But in countries where these habits are common. such as India and Ceylon. death rates from oral cancers are extremely high (Wynder. 1982). The links between tobacco use and a number of cancers come largely from retrospective and pro- spective epidemiologic studies. as well as from lab- oratory studies and animal experiments. Analytic 69 Risk Factors chemistry studies have shown that tobacco smoke consists of almost 4,000 compounds, many of which act as initiators. promoters, and cocar— cinogens. The major carcinogenic activity, though, comes from the tar. or particulate portion (Wynder, 7 1982). The tar contains carcinogenic agents. Chew- ing tobacco and snuff may contain large parts of N-nitrosamines that may be metabolized to car— cinogens. Viruses 70 Francis Peyton Rous of New York showed in I911 that a virus infection could increase the incidence of cancer in chickens, and in 1972, he was awarded the Nobel Prize for this research. Other studies have shown that viruses can increase cancer inci— dence in a wide spectrum of animal species. Al- though some viruses are associated with human cancers. other factors are believed to be responsi- ble for the development of cancer. Viruses may make cells more susceptible to the effects of radia- tion or chemical carcinogens, for example. Viruses may invade the genetic material ofa cell and affect the cell‘s protein production. causing . changes that can lead to uncontrolled growth. Un— der this theory. a whole virus need not be present ‘ to cause a change. A single gene. the basic unit of heredity that codes for a protein. may be enough to disrupt the cell’s normal function or make it more vulnerable to other carcinogens. Epidemiologic research has focused on the pos- sible roles of several viruses in human cancer. . Among them are Epstein-Barr virus (EBV), herpes simplex type 2 (HSVZ) and hepatitis B. Found throughout the world, EBV has been linked to Burkitt’s lymphoma. a cancer found in African children, and to nasopharyngeal cancer. which oc- curs mainly in adults in southern China and in parts of Southeast Asia. EBV also causes infec- tious mononucleosis, the benign “kissing disease" common among young adults in the United States. A second candidate, HSVZ, appears to be trans- mitted primarily by sexual activity and has been tentatively linked to cancer of the uterine cervix. But other factors are also important in the develop- ment of cervical cancer. V 1 The third candidate, hepatitis B virus, is linked to the high incidence of liver cancer in Africa, Tai- wan, and China. A chronic infection with the virus may be necessary for most cases of liver cancer to develop (Blumberg and London, 1982). Robert Gallo and coworkers at the National Can— cer Institute have found another type of virus in certain rare forms of human leukemia and lympho— ma (Poiesz, 1980) known as human T—cell leuke- mia-lymphoma virus, or HTLV. It infects human T- cells, white blood cells in the immune system, and is the first virus with genetic information coded as ribonucleic acid, or RNA, shown to play a role in human cancer. All other viruses linked with human cancer have genetic information coded as deoxy- ribonucleic acid, or DNA. The presence of antibodies for some viruses, which often indicates earlier viral infection. is linked to the high incidence of certain cancers. In Japan and in Caribbean countries, for example, 90 to 100 percent of T—cell leukemia or lymphoma pa— tients also have human T—cell leukemia-lymphoma antibody. And almost all Africans with Burkitt’s lymphoma have large amounts of antibody to Ep- stein-Barr virus. African children become infected with EBV when antibodies acquired from the mother disappear at 2 to 8 months of age (Biggar et al, 1978). Other factors acting with viruses are probably needed for cancer to develop. The necessary “cofactor” may be another virus, a hormone, an immune system deficiency, or an environmental factor. Cofactors are thought to be necessary be- cause only a small percentage of persons infected with these viruses ever seem to develop cancer. Thus, EBV is common in many areas that have very low incidences of the associated cancers, but the link between EBV and Burkitt’s lymphoma in Africa may require a predisposing factor like malaria. It is possible that vaccines will be developed against viruses linked to human cancers. An exist- ing vaccine against hepatitis B virus, for example, may reduce the risk of liver cancer. One approach to preventing cancers associated with viruses may be by reducing exposure to possible cofactors as they become identified. 7| Risks for Major Cancers Biliary Tract Biliary tract cancers are more common in other countries than in the United States where their in- cidence is low. In this country. American Indians and Hispanics are at greater risk than the general population. The biliary tract consists of the gallbladder. which stores bile made in the liver. and the ducts. which transport the bile to the small intestine. There the bile helps to digest dietary fats. Cancers may develop in the gallbladder or in the duct cells. Incidence increases with age: most cases of bili- ary tract cancer occur in persons over age 65. The epidemiology of cancers of the biliary tract closely resembles that of gallstones. but except for that as- sociation, little is known about the possible causes of these cancers. Gallbladder Cancer Cancer of the gallbladder is more common in wom- en than in men, perhaps due to hormonal changes during puberty and pregnancy that affect bile pro- duction. Gallbladder cancer risk rises with the number of pregnancies. In the United States. the incidence of gallbladder cancer is l.8 per 100,000 for women and l.() for men. Among American Indians in New Mexico. the incidence is 22.3 for women and 5.9 for men. Among Hispanics in New Mexico. incidence ranges from 9.2 for women to 2.5 for men (Young et al, 1981). Gallbladder cancer accounts for fewer than 1 percent of all cancers in the United States (Young et al. 1981). but it accounts for about 5 percent of all cancers among women in Latin American coun- tries (Albores-Saavedra et al. 1980). Incidence is also high in Israel, central and eastern Europe. and Japan. The lowest incidence is found among Indi- ans in Bombay (Waterhouse et al, 1982). Gallstones are the most important known risk factor for biliary tract cancer and especially for gallbladder cancer. Only a few of the millions of persons with gallstones ever develop biliary tract cancer, but about 70 percent of gallbladder cancer patients have also had gallstones (Fraumeni and Kantor. 1982). American Indians, with their very high risk for gallbladder cancer, also have a very 73 74 high incidence of gallstones. Controlling the risk factors for gallstones could decrease the incidence of both gallbladder and bile duct cancers (Fraumeni and Kantor, 1982). Biliary Duct Cancers Cancers of the bile ducts, especially in the ducts outside the liver. tend to occur more often among men than women and are less likely to be linked with gallstones. 1n the United States. incidence is 1.7 per 100,000 for men and 1.1 for women. But among American Indians in New Mexico. men have an incidence of 3.4 and women have an inci- dence of 6.1 (Young et al, 1981). In the past few decades. the incidence of biliary tract cancer in the United States has generally been decreasing among women and increasing among men (DeVesa and Silverman. 1978). The link between gallstones and bile duct cancer is not so strong as the link between gallstones and gallbladder cancer. Only about 30 percent of pa— tients with bile duct cancer also have gallstones (Fraumeni and Kantor. 1982). Other Biliary Tract Cancer Risks Several risk factors have been identified for general biliary tract cancer. Typhoid carriers. for example. have been found to have a high risk of developing biliary tract cancer. In some Asian countries. infec— tion with liver flukes appears to increase risk by causing inflammation of the biliary system. which may then lead to cancer. Patients with ulcerative colitis have about a 10 times greater risk of devel- oping biliary tract cancer than do healthy individu- als (Ritchie. 1974). Animal studies have shown that some ofthe chemicals used to process rubber are linked to bili- ary tract cancers. A study of workers in a rubber factory also found an excess of biliary tract can- cers (Mancuso and Brennan. 1970). Risks for Major Cancers Brain and Other Nervous System Cancers The nervous system consists of two anatomical parts: the central nervous system. which includes the brain and the spinal cord. and the peripheral nervous system. Cancers of the nervous system are of many different types and are classified ac- cording to tumor cell type and location. Nine out of 10 nervous system cancers are brain cancers and the most common type of brain cancer. ac— counting for more than half of all adult brain can- cers. is glioma, a fast-growing cancer usually located in the upper part of the brain. Brain cancers occur at varying rates according to sex. race and age. Overall. the incidence of brain and other nervous system cancers is 6.3 cases per 100.000 population per year among U.S. men. and 4.4 among women. The annual U.S. death rate from nervous system cancers is 4.9 per 100.000 for men and 3.2 for women (Young et al. 1981). Except for meningiomas. a benign tumor of the membranes that surround the brain and spinal cord. men have a higher incidence than do women of all types of benign and malignant nervous sys- tem tumors. These differences in incidence suggest a possible hormonal factor in the development of nervous system tumors. In the United States. brain cancers occur most often among whites. Among white men. they occur at a rate of 6.7 per 100.000. and among white wom- en. 4.6. Among blacks. they occur at a rate of 4.2 per 100.000 for men and 2.7 for women (Young et al. 1981). Whites develop gliomas more often than do blacks. while blacks develop meningiomas more often. Brain cancer is the second most common type of cancer in children after leukemia and occurs most often in children under 10. In adults. brain cancers occur most often between the ages of 55 and 79 (Young et al. 1981). Children have a higher incidence of medulloblas— toma. a cancer that affects the part of the brain connected to the spinal cord. It accounts for al- most a quarter of all childhood brain cancers. but fewer than 2 percent of adult brain cancers (Schoenberg. 1982). Little is known about the causes of brain can- cers. although studies have linked them with oc- cupational. environmental. viral and genetic factors. 75 76 Certain brain cancers appear to be more fre- quent among workers in particular industries. White men who worked in three oil refineries on the Texas Gulf Coast, for example. had a higher proportion of brain cancer deaths between 1943 and 1979 than did the general US. population (Thomas et al, 1982). Chemists (Olin and Ahlbom. 1980). pharmaceutical workers (Thomas and De- coufle. 1979). embalmers (Walrath. I983). and workers in rubber manufacturing plants who are exposed to vinyl chloride (Schoenberg, 1982) ap- pear to have more brain cancer than the general population. Cattle and sheep ranchers. dairy farm- ers, and grain millers have also been found to have a greater proportion of brain cancer (Milham, 1976). Long-term pesticide exposure of farm workers and of children raised on farms has been associ- ated with brain cancer development (Choi et al. 1970; Gold et al, I979). These studies link child— hood brain cancer with exposure to sick pets and farm animals. thus indicating a possible viral etiology (Schoenberg, I982). There is some clinical evidence that lead ex- posure may be linked to a type of glioma in chil- dren (Schreier et al, 1977), evidence supported by laboratory studies in which rats fed high-lead diets developed gliomas. More than 30 chemical com- pounds have been shown in animal studies to result in a high incidence of nervous system tumors and cancers. A few studies have shown possible genetic sus- ceptibility for some nervous system cancers. Reti- noblastoma. a rare eye cancer. is known to occur more often in families than among non-related peo- ple. as do certain gliomas (Schoenberg. I982). There is also a significant association between brain cancers in children and the presence of epi- lepsy in their siblings (Gold et al, 1979). Other factors that may be related to brain cancer include high-dose X-rays; consumption of sodium nitrate. a commonly used meat preservative; head trauma; and the use of barbiturates by pregnant women and by children (Schoenberg. I982). Brain cancers do not usually spread outside the central nervous system. but cancers from other sites may metastasize to the central nervous sys- Risks for Major Cancers tern. Breast cancer. for example. often spreads to the brain and spinal cord. Melanoma and cancers of the kidney and gastrointestinal tract may also metastasize to the brain (Schoenberg, 1982). Breast Breast cancer is the most common form of cancer (after skin cancer) among American women and accounts for more deaths among them than any other type. The high incidence rates of breast can- cer that prevail in the United States—85.6 per l0().- 000 for white women in 1973—77. and 72.0 for black women (Young et al. l981)—are also seen in other western and industrialized countries. including Canada. western Europe. Australia. New Zealand. and South Africa. Incidence in Asia. Latin Amer- ica. and Africa is low. and intermediate in eastern and southern Europe. In the United States. breast cancer occurs more often among women who live in cities than among those who live in rural areas (Blot et al. 1977). Though genetic factors may play some part in the variations in incidence seen in different locales. environmental factors also appear to be important. Migrant studies have shown that breast cancer inci- dence among women born in countries with low in- cidence increases when they move to the United States. where incidence is high. This has been seen among Japanese women who have moved to Hawaii and to the mainland United States (Haenszel and Kurihara. 1968). and among the daughters of European immigrants to the United States. Epidemiologic studies have helped to identify a number of factors that carry varying degrees of risk for breast cancer. To date. no one factor or combination of factors has been found that can be used to predict the occurrence of breast cancer in any one individual. Meanwhile. these are among the major risk factors that have so far been identi- fied: Age: In countries where the incidence of breast cancer is high. the rate increases sharply after age 30 and continues to rise. Among population groups with low or moderate incidence. there is often a 77 78 leveling off at the ages of menopause and after. For women in the United States, age is a major risk factor for breast cancer. Family history: Risk of breast cancer is in- creased when close relatives have had breast can- cer. A woman’s risk is increased twofold if her mother or sister have had it and sixfold if both have had it. Previous breast cancer: Women who have had cancer in one breast have a four— to fivefold risk of developing a second cancer (Kelsey. 1979). Reproductive experience: The risk for women who have children later in life increases with their age at first birth. Women who have never had chil- dren and women who have a first child after age 30 have a risk about three times greater than women who have a first child before age 18. Breast feeding does not seem to affect breast cancer risk. Menstrual history: Early onset of menstruation and late menopause both appear to increase breast cancer risk. Menopause induced by removal ofthe ovaries (usually in conjunction with hysterectomy) before age 40 reduces breast cancer risk consider— ably. Benign breast disease: Women with benign fibro- cystic breast disease confirmed by biopsy appear to have a three to five times greater risk of breast cancen Radiation: Large doses of radiation have been linked with the development of breast cancer both in studies of women exposed in Hiroshima and Nagasaki. and of women who were X—rayed many years ago for diagnosis of pulmonary tuberculosis. The much lower doses now used for chest X-rays and for mammograms, or breast X-rays. appear to carry little or no risk. Socioeconomic status: Women with high socio— economic status appear to be at greater risk of breast cancer but why this is so is not clear. Estrogens and oral contraceptives: There is some evidence for an increased risk of breast can- cer associated with the use of replacement es- trogens. particularly in long-term users. although the data are not all consistent (Thomas DB. 1982). Generally. the use of oral contraceptives has not been linked to increased risk of breast cancer (Brinton et al. 1982). even among women who took Risks (or Major Cancers the pill for more than ll years (CDC. 1983). One recent study found some evidence for an increased risk of breast cancer among women under age 36 who were long—term users of combination pills con- taining a high dose of progestin (Pike et al. 1983). Nutrition: There is some evidence that obesity is associated with increased breast cancer risk (de- Waard, I975), leading to speculation that high di- etary fat intake contributes to increased risk. To date, most of these associations have been derived from studies of geographical groups‘ and evidence on an individual basis is lacking. Overall. the evidence suggests that hormonal factors may play a part in breast cancer in addition to genetic predisposition. This hypothesis would account for the higher risk seen in women with a long menstrual history. because of the long ex- posure of breast tissue to hormone stimulus. Large-scale studies that would take combinations of risk factors into account have not yet been done. Childhood The incidence of cancer among children under age 15 in this country each year is 12.4 per 100.000 among whites and 9.8 among blacks. Though low compared with the incidence of some adult can- cers. cancer is second only to accidents as the leading cause of death among children (Young et al. 1978: Young and Miller. 1975). More than 40 percent of childhood cancers occur in the very young—age 4 and under. More encouraging is the fact that five-year rela- tive survival rates for many of the childhood can- cers have increased dramatically in this country from the 19605 to the |970$ (Myers and Hankey. 1980). The five-year survival rate for acute lympho- cytic leukemia in children. for example. has in- creased from 1 percent to almost 50 percent in some groups of patients (Young et al. I978). The incidence of the childhood cancers varies greatly throughout the world, depending on the type. The acute leukemias account for about 85 percent of them in this country. But in tropical Af- rica. Burkitt‘s lymphoma. a cancer of the lymph 79 80 system. accounts for more than half ofchildhood cancers. In Ankara. Turkey. almost half are acute myelomonocytic leukemia (AMML). but AMML accounts for only 4 percent ofchildhood cancers in this country. There are also varying age trends for the child- hood leukemias. Among white Americans. western Europeans. and. more recently, Japanese, there is a peak in cancer incidence at ages 2 through 4 (Mil- ler. 1977). Among black children in the United States there is no such age peak. These are the major chilhood cancers in this country: Acute leukemias Acute leukemias are the most frequent. and acute lymphocytic leukemia (ALL) accounts for most of these. The incidence is slightly higher among boys than among girls. About 40 percent of acute leuke- mia cases in children have been linked with chro- mosome disorders. Ionizing radiation—energetic rays that cause molecules to gain or lose electrons—can cause leu- kemia. When warnings about radiation were widely publicized to the medical community in the mid-l950s. safety procedures for diagnostic X-rays were tightened and. by the l960s, leukemia inci— dence had fallen among all groups under age 75. Studies of the possible effects of prenatal irradia- tion have not been conclusive (Li. 1982). Some anticancer drugs and at least one industrial solvent. benzene. can cause leukemia in adults. but few drugs and chemicals have been shown to in— duce cancer in children (Hoover and Fraumeni. I975). Lymphomas These cancers ofthe lymph system are the third most frequent and Hodgkin‘s disease accounts for about half ofthem (Young et al. [978). Hodgkin‘s disease is rare in early childhood. peaking in fre- quency among young adults. Some studies have linked its occurrence to high socioeconomic status. but others have not found this link (Li. I982). There may be an increased risk of developing Hodgkin‘s disease among siblings of those who get it. particularly among brothers. The non-Hodgkin‘s lymphomas. or NHL. are also more common among boys and occasionally cluster in families. Like leukemias. they are linked with several rare. genetically determined. immune system diseases. Further evidence for the link of NHL with immune system disorders comes from the observations that persons who receive kidney transplants. and have thus been deliberately immu- nosuppressed. have a 50-fold increased risk of NHL (Fraumeni and Hoover. 1977). Burkitt‘s lymphoma. the cancer common among African children. has been linked with Epstein— Barr virus. or EBV: raised levels of antibodies to the virus—indicating that an individual has been exposed to it—have been found in many children with the lymphoma. A cause-and-effect rela— tionship has not been shown. Central nervous system Cancers of the central nervous system—the brain and spinal cord—account for about a fifth of can- cers in children under age 15 and tend to occur in the first 10 years of life. A number of genetic disor— ders seem to be linked with excess risk of develop— ing these cancers. Bone cancers These account for about 5 percent of childhood cancers. The incidence of osteosarcoma. the most common bone cancer. peaks in late adolescence after a period of rapid bone growth (Glass and Fraumeni. 1970). It usually develops in the weight- bearing bones of the legs and pelvis. and particu- larly in the bone areas where most growth takes place. Ewing‘s sarcoma. another form of bone can- cer. accounts for about a third of bone cancers in US. white children. but is rare in black children anywhere. 8| Soft tissue sarcomas Ofthe soft tissue sarcomas. rhabdomyosarcoma is the most common. This cancer shows two distinct age peaks—one before age 5 and the other in the teens (Li and Fraumeni. l969(a)). Rhabdomyosar— comas of the head. neck. and genitourinary system form the first peak. and there is some suspicion that these cancers form before birth. Genetic factors may play a role in the develop— ment of sarcomas of bone and soft tissue. There are several genetic disorders that predispose to sar- comas. and some of them also occur as compo- nents of family cancer syndromes (Li and Fraumeni. 1969(b); Blattner et al. I979). Osteosar- coma tends to develop in some genetically caused bone lesions (Glass and Fraumeni. 1970). Because Ewings’s sarcoma is seen so rarely among black children. genetic factors may play a part in its development. also (Miller. 1977). Neuroblastoma This cancer arises in one type ofimmature nerve cells. and accounts for more than 5 percent of childhood cancers in both black and white chil— dren. The cancer may be present at birth. and more than half of all cases appear before age 2 (Miller et al. 1968). Neuroblastomas appear to arise during fetal life. and may be linked with inborn de- fects of nerve tissues (Knudson and Meadows. I976). Wilms’ tumor A cancer ofthe kidney. Wilms‘ tumor is usually di- agnosed between ages I and 3 (Young and Miller. l975). About 14 percent of Wilms’ tumor patients also have congenital anomalies. among them aniridia (missing iris ofthe eye). Children who have aniridia and the tumor consistently lack a segment of a chromosome (number 13). But in one set of twins. both lacked an iris and both lacked the chro- mosome segment. yet only one of the twins devel- oped Wilms’ tumor. This finding lends support to the idea that the predisposition to the cancer is in— herited. but that some second event is needed for the cancer to develop. Risks for Major Cancers Retinoblastoma About 40 percent of cases of retinoblastoma. an eye cancer. are inherited. Like Wilms‘ tumor. 21 small percentage of these eye cancers has been linked to a missing segment of chromosome. Colon and Rectum The colon and rectum together make up the large bowel. or large intestine. The colon refers to the upper five or six feet of the large intestine; the rec- tum to the last five or six inches. Risk factors vary for the two cancers, as do their patterns of inci- dence. but the two are often lumped together as “colorectal cancers“ to compare death rates over time and among different countries. Colorectal cancers caused an estimated 58.000 deaths in this country in 1983 (ACS. I983). second only to lung cancer. Five-year survival for both cancers has improved markedly over the past 20 years. but is still under 50 percent. The incidence of colorectal cancers varies wide- ly throughout the world. with up to twentyfold dif- ferences. In 1970. for example. the age-adjusted in- cidence of colon cancer among men was 30.1 per 100.000 in the State of Connecticut. but only 1.3 in lbadan. Nigeria (Waterhouse et al. I976). High in- cidence and mortality rates are seen in the highly developed countries of North America. northern and western Europe. and New Zealand. The lowest rates are seen in Asia, Africa. and most ofthe Lat- in American countries (Logan. 1976). In general. in areas where the incidence ofcolo- rectal cancers is high. colon cancers occur about twice as often as rectal cancers (Schottenfeld and Winawer. 1982). This is true in the United States. for example. In areas where the incidence of both cancers is low. the two occur with about equal fre- quency. Wide variations in death rates from colorectal cancers are found within the United States. The death rates are highest for men and women. blacks and whites. in the Northeast. particularly in New Jersey. Massachusetts. southern New York State. and urban areas around the Great Lakes; death rates are markedly lower than the US. average in 83 84 parts ofthe South and Southwest (Mason et al. 1975). Studies have also shown correlations of colon cancer death rates with the degree of urbani- zation and socioeconomic status (Blot et al. 1976). The incidence of colon cancer in this country has been rising slightly for white men and women and for black women over the past two decades (Young et al, I981) and has increased sharply for black men. The incidence of rectal cancer has in- creased for white men. but has remained almost constant for white women and blacks of both sexes. Among the risk factors for colorectal cancers are age. family history. ethnic background. history of inflammatory bowel disease. and degree of urbani- zation. In recent years. diet has come under in— creasing scrutiny as a major risk factor. Age: The incidence of colon and rectal cancers increases slowly after adolescence and rises sharp- ly after age 50. In the 55 to 59 age group, the aver« age incidence per |0().000 for all races and sexes combined during 1973—77 (Young et al. l98l) was 60.2 for colon cancer and 32.7 for rectal cancer. By age 75 to 79. the incidence of colon cancer rose to 285.6 and that of rectal cancer to 104.7. Family history: As with many cancers. subsets of colorectal cancer cases appear to be largely due to familial background. For example. individuals with familial polyposis—a family tendency to de— velop colon polyps—have an extraordinarily high risk of developing colon cancer: so high. in fact. that surgery to remove the colon is recommended as a prevention measure. There are also families in whom certain groups of cancers develop more fre- quently than in the general population; colorectal cancers tend to occur in some ofthese groups. Ethnic background: A number of U.S. studies have pointed up disparities in colorectal cancer in- cidence and death rates according to ethnic back- ground. One recent case-control study in rural Ne- braska. for example. showed an increased risk of colorectal cancers among persons of Czech back- ground (Pickle et al. I984). It has often been diff- cult with these ethnic groups to separate genetic factors from environmental factors like diet. Mi- grant studies may have provided more valuable clues. A number of these migrant studies have shown that risks for colorectal cancers change as ethnic groups move from one country to another. The death rates for colorectal cancer in Japan. for ex- ample. are far lower than in the United States. But when Japanese move to this country. their risk of developing colon cancer increases markedly (Schottenfeld and Winawer. 1982). Likewise. the incidence of colorectal cancers in Poland is far lower than in the United States. but in one study (Staszewski. 1972). the death rates for colorectal cancers among Polish-born immigrants to the United States were similar to the US. population. These and other studies indicate that migrant popu- lations take on the colorectal cancer rates of their new country. Moreover. this seems to happen quite quickly. within one generation. This phenomenon is also being studied in “migrant" populations within the United States. When urban Northerners migrate to the South. where colorectal cancer death rates are lower. the rates remain low. suggesting that the migrants as- sume the new. lower risk and do so in a fairly brief period of time. The National Cancer Institute is conducting a study in Florida to see what might ac- count for this. History of inflammatory bowel disease: Individu- als who have had extensive ulcerative colitis for ID years or more have a risk of developing colorectal cancer 10 times greater than the general popula— tion. There is also evidence that persons with Crohn’s disease. a chronic inflammation of the in- testine. have a heightened risk of developing colo— rectal cancers. Both genetic and environmental factors are believed to be responsible for these two diseases (Kirsner, 1980). Degree of urbanization: As noted above. colo— rectal cancers occur more often in urban. indus- trialized countries than in rural countries (Japan is an exception). This differential lends more support to the importance of environmental or lifestyle fac— tors for these cancers. but what those factors are is not clear. Diet: In general, high dietary intake of fats—in meat. milk and milk products. salad and cooking oils. and margarine—is linked to an increased inci- dence of colorectal cancers. particularly colon can- 85 86 cer, and high intake of dietary fiber—in fruits, veg- etables, legumes, and whole grain products— appears to protect against colorectal cancers. par- ticularly colon cancer. The evidence is not always consistent. however. International studies that correlate fat intake with colon cancer are more consistent (Armstrong and Doll, 1975; Knox, 1977) than studies of various groups within one country. A study of groups in different parts of Great Britain (Bingham et al, I979), for example, found no strong association be- tween fat intake and colorectal cancer death rates. And a study of Mormons in Utah, who have less colorectal cancer than other Americans (Lyon and Sorenson, I978), also failed to show a link; Mor- mons tend to eat more beef than do other Ameri- cans. Some case-control studies have shown strong links between fat intake and colon cancer, and others have failed to demonstrate this link (Miller et al, 1983). Dietary fiber intake may account for some ofthe conflicting data about dietary fat, if fiber has a pro— tective effect even when fat intake is high. Intake of milk fats was found to be similar, for example. for groups in rural Finland and in Copenhagen, Denmark, but colon cancer incidence was much lower among the Finns, who consume large amounts of high-fiber, unrefined rye bread (Mac— Lennan et al. 1978). lffiber protects against colo- rectal cancers, it may do so by speeding the transit time of fecal matter through the bowel, thus de- creasing the time that carcinogens in the fecal mat— ter are in contact with the bowel wall. Or. fiber may inhibit production of fecal mutagens (Erich, I979). The American diet typically includes more than 40 percent of total calories in fats. In Japan, where the incidence of colorectal cancers is far lower. fats make up only about 20 percent ofthe total calories in the diet. Some current dietary guidelines in this country are recommending that fat intake be de— creased to a level of about 30 percent (Palmer and Bakshi, I983). Diets high in fruits and vegetables appear to pro- tect against colorectal cancers. A cohort study in Minnesota (Bjelke, 1978) demonstrated this protec- tive effect. It has also been shown in one study Risks for Major Cancers that the cruciferous vegetables—cabbage. broccoli. Brussels sprouts. cauliflower—exert a protective effect against colorectal cancers (Graham et al. 1978). What in vegetables protects against colorectal cancers is not clear—vitamins A or C, other micro- nutrients, fiber, or the substances found in the cab- bage family. One case-control study of vitamin C consumption, for example, showed no association with colon cancer (Jain et a1. 1980). Esophagus Cancer of the esophagus is fairly rare in this coun- try with an overall incidence—men and women. all races combined—0f 3.6 cases per 100.000 popula- tion. It has remained at about this level for the past 40 years. Even with the best medical treatment. though. cancer of the esophagus is almost invari- ably fatal. There is strong evidence that esophageal cancer is caused by environmental. or lifestyle. factors. The chief indication for this is the very sharp geo- graphic variation in incidence. with “pockets“ seen in different locales (Day and Munoz, 1982). In one province in the Kazakhstan region ofthe USSR east of the Caspian Sea. the incidence soars to greater than 130 per 100,000 for both men and women. High rates are seen also in parts of China. northern Iran, France, and southern Africa. In most other parts of the world. the incidence ranges from 2 to 20 cases per 100,000. and occurs most often among men. In the United States, the inci- dence ranges from a high of 28.6 among black men in Washington, DC. to a low of 1.9 among Ameri- can Indian men in New Mexico. If the disease is caused by lifestyle factors, it might. therefore. yield to prevention. Although the causes for all of the clusters are not known, major risk factors have been identified among some pop- ulations. In parts of the United States, in Denmark. and in France. case-control studies have found that heavy consumption of alcohol and tobacco are major risk factors for esophageal cancer. A study conducted in the French province of Brittany (Tuyns et al, 87 88 1977) shows synergistic, or combined. effects of to- bacco use and alcohol intake. In Linhsien County. People‘s Republic of China, a prime suspect for esophageal cancer is a pickled vegetable mix pecu- liar to that area (Miller, 1978). More recently. a case-control study of black men in Washington, DC. (Pottern et al. I981). which has a higher incidence than any other US. city. showed that excess alcohol consumption was the major risk factor. accounting for about 80 percent of esophageal cancers. Poor nutrition was also im- plicated as a risk factor in this population. In Bombay (Jussawalla, I971). alcohol alone does not explain the risk of esophageal cancer. Smoking the local cigarette (“bidi”). coupled with drinking local brew, does seem to account for the risk in men, but not in women. In northern Iran and in northern Afghanistan, where the incidence of esophageal cancer is high for both sexes, both men and women follow the practice of eating the residue from opium pipes (Gobar, 1976). A number of other promising leads are being fol- lowed in studies of populations with esophageal cancer. Low socioeconomic status and poor nutri- tion—particularly diets deficient in riboflavin, vi— tamin C, and vitamin A—appear to be linked to in- creased risk of esophageal cancer in some areas. Esophageal cancer has also been linked with sid— eropenic dysphagia, an inflammation of the esophagus coupled with iron deficiency. Zinc defi— ciency has also been implicated as a risk factor in case-control studies of esophageal cancer. Most re- cently. vitamin A deficiency was found linked to esophageal cancer risk in a study of American men by Mettlin and Graham (1981). Risks for Major Cancers Hodgkin’s Disease Hodgkin‘s disease—a form of lymphoma, or cancer of the lymph system—is fairly rare in this country. with an incidence of 3.5 cases per 100.000 popula- tion per year among men and 2.4 among women (Young et al. 1981). It is still one ofthe most com- mon cancers among young adults. however. And while extremely rare in US. children under age 5. it is often one of the most common cancers of childhood in less developed countries (Correa and O‘Conor, 1971). Its incidence is characterized by two age peaks—one in the early 20s and another. higher peak in the late 705 (Young et al. 1981). This pat- tern prevails throughout the world except in Japan. where only the second peak is seen. Because of this pattern of occurrence of Hodgkin‘s disease. there is some suspicion that there may be two. or even three. different disease processes at work (Grufferman. 1982). Despite its rarity. Hodgkin's has received a great deal of attention from researchers. One reason is that it has yielded so well to therapy injust a dec- ade. Overall. more than f0u1‘<fifths of patients un- der age 35 are now living 5 years or more after treatment (Ries et al. 1983). Among patients whose disease is found early. the 5-year survival rate is even higher. Hodgkin‘s disease has also interested research- ers because of a persistent notion that it might be caused by an infection. These are among the etiologic leads that have been followed: Clusters Two apparent geographic clusters of Hodgkin‘s dis- ease were identified among high school students in Albany, N.Y.. and in one Long Island community (Vianna et al. 1971; 1973). The studies concluded that students and teachers at the schools were at increased risk of developing the disease because a number of cases had occurred among students. The studies implied that the schools might serve as a focal point for spread of the disease or that con— tact with patients can spread it. Subsequently. though. a series of studies per- formed in Boston; Oxford. England; Connecticut. and Atlanta were unable to confirm the two earlier studies. 89 90 Doctors and nurses Another study in upstate New York (Vianna et al. I974) indicated that physicians had a greater than normal risk of developing Hodgkin’s disease, pre- sumably because of their greater exposure to pa- tients. Later studies of British physicians (Smith et al. I974), four groups of US. medical specialists (Matanoski et al, 1975), and doctors and nurses in Boston (Grufferman et a], 1976) showed that these health-care providers were not at greater risk than the general population. Mononucleosis. Epstein—Barr virus (EBV), a herpes virus, has been isolated from patients with Burkitt’s lymphoma. the most common cancer among African children. EBV is also linked with one type ofinfectious mononucleosis. known among college students as “the kissing disease." These observations led to a search for associations between mononucleosis and Hodgkin’s disease. One of the first studies (Miller and Beebe, I973) looked at records of more than 2,000 veterans who had had infectious mononucleosis but found no convincing evidence of increased risk of Hodgkin‘s disease. Subsequent studies of other groups were also inconclusive, although a Danish study (Rosdahl et al, 1974) showed a twofold excess of Hodgkin’s disease after mononucleosis. Other Possible associations between Hodgkin’s disease and tonsillectomy, polio‘ multiple sclerosis, and some other conditions have also been explored, but without success. Thus. no significant environmental risk factors have been identified for Hodgkin‘s disease. though there is an association between it and educational level: it appears to occur excessively among the well—educated. As for host factors. ataxia telangiec- tasia, a rare, genetic immune deficiency disease. is the only genetic condition that appears to pre— dispose individuals to Hodgkin‘s disease. Risks for Major Cancers Leukemia Leukemia is the term for a variety of cancers that arise in blood and bone marrow cells. The various leukemias, though, are quite different from each other in terms of cause, therapy, and outlook, de- pending on which cells they affect. Chronic lym— phocytic leukemia, for example, seems more akin in some ways to lymphoma—cancer of the lymph system—than to other leukemias. There are four main types: acute lymphocytic leukemia (ALL), acute myelocytic leukemia (AML), chronic myelocytic leukemia (CML), and chronic lymphocytic leukemia (CLL). Together they account for about 5 percent of the total annual cancer incidence in this country. They account for about 45 percent of cancers in children. Geographic patterns in leukemia incidence show some marked variations that may have more to do with ethnic factors than with geography. Leukemia incidence is lower among children in Africa and Ja- pan than in this country, and CLL is very rare in Japan and China. It is equally rare among these Asians living in other countries. And one rare type of leukemia, acute myelomonocytic (AMML). ac— counts for 40 percent of childhood leukemias in Ankara, Turkey, compared with only 4 percent in the United States. In the United States, there is slightly more leu- kemia among whites than among blacks, and Jews have a somewhat higher incidence than other whites. The male-to-female ratio overall is about 1.7 to 1, and men have almost two times more CLL than do women (Young et a1, 1981). There are wide age variations in leukemia inci- dence in this country. The overall annual incidence ranges from about 2 cases per 100,000 in young adults to 30 or more per 100,000 over age 65. There is also a peak among white children of almost 7 cases per 100,000 under age 5. The specific forms of leukemia are very different in their age patterns. Acute leukemias are common at all ages. accounting for almost all leukemias in children and young adults and for about two—fifths ofthose in older adults. In young children. these are usually acute lymphocytic leukemias (ALL), but after puberty most are acute myelocytic leuke- mias (AM L). CLL is almost never seen before adulthood but is common over age 50. 91 The leukemias appear to be caused by a number of factors that act alone or in combination (Miller. 1979). Some are host factors. such as genetic traits and immune status; others are environmental fac- tors such as radiation. chemicals. or viruses. A number of family “clusters“ of leukemia have been investigated at various times in this country. but few. so far, have shown clear evidence of a fam- ily predisposition for the disease. The near-absence of CLL among Orientals. on the other hand. ar- gues for a genetic resistance to it among them. Leukemias are associated in some instances with abnormal chromosome patterns. such as the Phila- delphia chromosome (characterized by a transloca- tion. or swap. of parts between two chromosomes) found in the bone marrow cells of persons with CML. Leukemia is also about 20 times more com- mon than usual in persons with Down‘s syndrome. And it occurs more often in persons with rare con- ditions—some of them hereditary—that are charac- terized by increased chromosome breakage. It has been known for some time that leukemia can be caused by ionizing radiation. This is radia- tion characteristic of very energetic rays—like X- rays or gamma rays— that can cause molecules to gain or lose electrons. Some of the first evidence of the leukemia—causing potential of radiation came from the effects seen on radiation workers soon after the discovery of X-rays. For some time. US. radiologists had elevated rates of leukemia. but their risk has decreased with improved safety tech- niques. (These include better shielding. lower doses, and more concentrated fields of exposure.) The chief sources of ionizing radiation are X-ray therapy and diagnostic X-rays. Studies of Japanese who lived through the atomic bombs (Beebe et al. 1978) at Hiroshima and Nagasaki showed that all forms of leukemia except CLL increased in inci- dence with a peak about 5 years later. The lag time between exposure and the development of leuke- mia varied according to the type of leukemia and the age at exposure. Evidence for X-ray therapy as a cause of leuke— mia has come from a number of studies of persons who received such therapy some years ago. Among the earliest was a British study of men who received X-ray therapy for a rheumatic inflamma- tion of the spine (MRC. 1956). Evidence that diagnostic irradiation could cause leukemia came from followup studies of individuals who received the radioisotope thorium dioxide as an X-ray contrast agent during the years l928—l955. and later showed increased rates of leu- kemia (da Silva Horta et al. 1974) and liver cancer. In I956. the National Academy of Sciences-Na- tional Research Council in the United States (NAS-NRC. 1956) and the Medical Research Council in Great Britain (MRC. 1956) issued re- ports on the biological effects of ionizing radiation. Both cautioned against the use of routine fluor- oscopy (an X-ray technique) for fitting children‘s shoes and against other nonessential uses of heavy radiation. There was wide publicity about these re- ports. and by 1960. declines in leukemia incidence and mortality for all age groups could be seen in Great Britain. the United States. and Norway. Increased leukemia incidence has also been re- ported among military personnel present at nuclear tests and among populations in some areas that re- ceived fallout from such tests. A possible link with radiation is still being studied. A number of chemicals used in the workplace are linked with increased risks ofleukemia. Among them are benzene and other related solvents used in industry. Some of the anticancer drugs have also been linked in recent years with an increased risk of leukemia. It has been thought for some time that infectious agents might play a part in human leukemia: vi- ruses do cause leukemias in cats. chickens. mice. monkeys. and cattle. Epstein-Barr virus (EBV). which has been linked to one type of lymphoma. or cancer ofthe lymph system. was one of the chief suspects. Extensive studies were not able to show a link between either animal viruses or EBV virus and leukemia in humans. Recently. though. a virus was found that appears to cause an unusual form of leukemia and lympho- main humans (Poiesz et al. I980). Antibodies to this virus. the human T-cell leukemia/lymphoma virus. or HTLV. have since been found among per- sons living in parts of Japan. the Caribbean basin. Africa. some other locales. and sporadically in the United States and Western Europe. (Finding anti- 93 Risks for Major Cancers bodies in these individuals indicates that they have been exposed to the virus.) This finding raises the hope that the forms of leukemia or lymphoma linked with this virus may one day be prevented. and that new understanding of other human leuke— mias may come from these studies. Liver 94 Primary liver cancer is cancer that first develops in the liver and may then spread to other organs. Al- though it accounts for only about 0.6 percent of all diagnosed cancers in the United States, it is a lead- ing cause of death in other parts of the world (Young et al, 1981). There has been a steady decrease in deaths from liver cancer in the United States over the past 30 years (American Cancer Society, 1982), and pri— mary liver cancer is now fairly uncommon in many Western countries. Incidence generally increases with age, except for a small peak during childhood. In the United States, it is highest in the elderly. Be- tween 1973 and 1977, about 21 per 100,000 white men age 80 to 84 developed liver cancer each year. The incidence for black men age 75 to 79 was about 33 cases per 100,000 each year (Young et al, 1981 ). Incidence is also higher in black women than in white women. The highest incidence in the world has been found in Mozambique, Africa. Between 1956 and 1961, there were some 110 cases of primary liver cancer per 100,000 males per year, and about 29 cases per 100,000 females per year (Prates and Torres, 1965). Incidence is also very high in Hong Kong, Taiwan, and Singapore, and among some Chinese-Americans (Waterhouse, 1976; Szmuness, 1978). The most important factor in the occurrence of liver cancer throughout the world may be hepatitis B virus. In some African populations, more than 90 percent of patients with liver cancer also have a chronic hepatitis B virus infection (Kew et al, 1979). In Taiwan, where primary liver cancer ac- counts for 20 percent of all cancers, persons in— fected with hepatitis B virus often develop cir— rhosis and are at very high risk for liver cancer (Beasley, 1978; Beasley and Lin, 1981). Research- ers hope that a recently developed vaccine against hepatitis B infection will help prevent liver cancer in high-risk groups. Men are more likely to develop primary liver cancer than women, possibly reflecting their higher exposure to risks factors. The male hormone tes- tosterone increases the incidence of liver tumors in laboratory animals and may be a factor in humans (Falk, 1982). Alcohol consumption is linked with cirrhosis, a liver disease, and cirrhosis is linked with liver can— cer. Sixty to 90 percent of liver cancer occurs in association with cirrhosis but the relationship be- tween the two diseases is not completely under- stood (Falk, 1982). Angiosarcoma is a rare type of primary liver cancer linked to workplace exposure to vinyl chlo- ride during the production of plastics. This type of cancer is so rare in the general population that even a few cases among exposed workers yielded evidence that vinyl chloride is carcinogenic. Between about 1930 and I955, Thorotrast was used as a radioactive contrast agent in diagnostic imaging. lt accumulates in the liver, exposing the organ to excess radiation, and causing an increased risk of angiosarcoma. A widely distributed fungus that infests poorly stored peanuts and other foods produces aflatox- ins, some of the most carcinogenic agents known. Aflatoxins cause liver cancer and other types of cancer in laboratory animals and may, in conjunc- tion with hepatitis, be a factor in the high rates of primary liver cancer in Africa and Asia. In the United States, the Food and Drug Administration controls the amount of aflatoxin allowed in peanut butter, but there have been cases of tainted animal feed and fish meal (Sinnhuber et al, 1968). Im- proved food handling and storage in areas where liver cancer is common could result in decreased incidence rates. 95 Risks for Major Cancers Lung and Larynx 96 Primary lung cancer in the United States accounts for about 14 percent of all cancer cases (21 percent in males and 7 percent in females). And because of its high death rate. it accounts for 23 percent ofall cancer deaths—3| percent in males and 12 percent in females (Young et al. 1981). The incidence of lung cancer is increasing rapidly in most areas ofthe world. and is already the leading cause of cancer in many countries. The annual age-adjusted incidence among men exceeds 100 per 100.000 population in Great Britain. U.S. blacks. and Finland. The incidence is much lower in China. India. and a number of African and Latin American countries. where rates are less than 25 per 100.000 population. Lung cancer still ranks as one of the three most common cancers in males throughout the world. and probably now outranks stomach cancer as the most common cancer in males (Fraumeni and Blot. I982). The incidence is generally lower in females. whose rates are usually less than 25 per 100.000. This difference is thought to be due to differences in tobacco consumption. In the United States. lung cancer incidence has risen more sharply in females than in males in re- cent years. reflecting the growing popularity of cig- arette smoking among females over the past sev- eral decades. The lung cancer death rate among women in this country is expected to surpass that of breast cancer is a few years (Horm and Asire. [982). Cigarette smoking is the major cause of lung can- cer and is estimated to cause 85 percent of lung cancer deaths. Lung cancer mortality increases with increasing doses. as determined by number of cigarettes smoked daily. smoking history. and inha- lation patterns. Those who smoke two or more packs a day have death rates 15 to 25 times greater than nonsmokers. Cessation of smoking. though. reduces the risk of death from lung cancer: after 15 years. former smokers have lung cancer death rates only about two times greater than non- smokers. Those who smoke filtered. low-tar ciga- rettes gain some benefit. but they still have death rates much higher than nonsmokers. The link between cigarette smoking and lung cancer was first suspected in the l920s and 19305. and was supported in the l950s by case-control studies in the United States and Great Britain. The evidence became conclusive through surveys of large cohorts of smokers. Two such studies were the 6-year American Cancer Society study of more than a million persons (Hammond. 1972) and the 20-year study of 34.000 British. male physicians (Doll and Peto. 1976). Both surveys showed an ele- vated relative risk of lung cancer among smokers. Studies of occupational groups have identified several other respiratory carcinogens. although it is difficult to assess their precise impact. It has been found that tobacco smoke interacts with some of these occupational carcinogens. such as asbestos and radon. Exposure to airborne asbestos appears to exert the largest cancer threat in the workplace. raising the risk of lung cancer and mesothelioma (a cancer that arises in the lining of the chest cavity. or mesothelium) as well as asbestosis. a lung disease (Fraumeni and Blot. 1982). Epidemiologic studies over the past 25 years have indicated that the risk of developing these three diseases is substantially higher for workers in a number of asbestos indus- tries. including miners and millers. and textile. in- sulation. shipyard and cement workers. Lung can- cer is the major asbestos-related disease. and accounts for death in about 20 percent of some men exposed to asbestos for long periods of their work life (Selikoffet al. 1979). Even men who worked for short periods in shipyards during World War II have a higher risk of developing lung cancer than workers never exposed to asbestos (Blot et al. I978. 1980). Asbestos-induced lung cancer is characterized by a lag time of 20 or more years between ex- posure and onset of disease. It is also evident that asbestos works synergistically with cigarette smok- ing; a study of lung cancer deaths among insulation workers (Hammond et al. 1979) showed that risk was increased fivefold from asbestos exposure alone. tenfold from smoking alone. and fiftyfold from combined exposure. Thus. reducing exposure to either agent could be expected to reduce risk. An increased risk of lung cancer has been estab- 97 98 lished among uranium miners in the Colorado Plateau, related to inhalation of radon daughters. (Radon daughters are decomposition products of the gaseous element radon. formed by the disin- tegration of radium. Radium occurs in small amounts in pitchblende and other uranium miner- als.) Radon is probably responsible also for the el- evated risk of lung cancer among hardrock miners in this country (Lundin et al. l97l). As with as- bestos. the effects of radon seem to be potentiated by cigarette smoking. Lung cancer is also one of the major effects of high doses of ionizing radiation. This is radiation produced by very energetic rays—like X-rays or gamma rays—that can cause molecules to gain or lose atoms. Besides the increased risk among ura- nium miners exposed to alpha radiation from radon daughters. excesses of lung cancer have been re- ported among some patients who received radia— tion therapy. and among atomic bomb survivors in Japan (Beebe et al. 1979). where both gamma rays and neutrons were released. A number of other occupational agents contrib- ute to the incidence of lung cancer. among them mustard gas. chloromethyl ethers. chromium. nick- el, and inorganic arsenic. Air pollution has been suspected as a cause of lung cancer. but it has been difficult to establish definite links. Major problems arise when trying to measure low-level effects of any carcinogen in a general population. Finally. there is growing evidence of an in- creased incidence of lung cancer among persons with diets low in vitamin A. particularly when combined with heavy smoking (Mettlin et al. 1979). This finding agrees with laboratory studies showing that vitamin A. and its chemical cousins. as well as beta-carotene. a precursor of vitamin A. protect against the induction and progression of tumors arising in various organs (Sporn. 1977). Cancer of the larynx, or voicebox. has an inci- dence pattern like cancers of the mouth and throat: it occurs more often among men than among wom- en, and more often among blacks than among whites. The annual incidence of laryngeal cancer among US. white men is 8.3 cases per 100.000 population. and [.3 among white women. Among Risks for Major Cancers black men. the annual incidence is 12.1 per 100.000. and 1.9 among black women (Young et al. 1981). There is some evidence that cancer of the larynx is increasing. especially among women. in more developed countries (Austin. 1982). In the United States. the median age of onset for laryngeal can- cer is 62 among whites and 58 among blacks (Young et al. 1981). Risk factors for laryngeal cancer include tobac- co. alcohol. asbestos. and nickel and mustard gas exposure (Austin. 1982). As with cancers of the lung. mouth and throat. many cases of laryngeal cancer can be attributed to cigarette smoking. Cig- arette smokers have almost a tenfold greater risk for laryngeal cancer than do nonsmokers. and risk increases with increased cigarette smoking (Wynder et al. 1976). Heavy alcohol consumption is also a risk factor. and tobacco and alcohol to— gether appear to act synergistically. Melanoma Melanoma is a cancer of melanocytes. the skin cells that produce the dark pigment melanin. Melanomas can occur almost anywhere on the body. but in light-skinned people they occur most often on the trunk in men and on the lower legs in women. The head. neck. and arms are other com- mon sites. In dark-skinned people. melanomas oc- cur most often on the palms ofthe hands and the soles of the feet. Worldwide. whites have the highest incidence of melanoma; it is far less frequent among Africans. Polynesians. and Asians (Waterhouse et al. 1982). In this country. the annual incidence is 6.7 per 100.000 among white men and 6.0 among white women. compared with 0.6 for black men and 0.7 for black women (Young et al. 1981). In Australia. the incidence is close to 20 per 100.000 among white men and women (Waterhouse et al. 1982). And worldwide. the incidence has been rising sharply. doubling every decade over the past 30 years. Melanoma is related to exposure to ultraviolet (UV) radiation. but not so directly as are the more 99 l ()0 common nonmelanoma skin cancers. basal and squamous cell. There is some evidence that the nonmelanoma skin cancers are related to cumulative exposure to UV radiation (Scotto et al. 1975). while melanoma is related to heavy. blister- ing overdoses. But other factors also seem to play a part in the development of melanoma. One such factor is familial predisposition: a re- cent genetic study suggested that melanoma may occur as an autosomal dominant trait (Greene et al. 1983). Other recent studies have also identified the dysplastic nevus syndrome as a melanoma precur- sor among melanoma families. A nevus is a true mole. made up of a cluster of melanocytes. and is not be confused with other pigmented lesions of the skin such as seborrheic and actinic keratoses. freckles. and other “spots.“ (The darker color of these other lesions. as well as the tan that results from sun exposure. result from transfer of melanin from melanocytes to non-pigment-forming cells.) Normally benign. the nevus may undergo abnor- mal changes. or dysplasia. which may then pro- ceed to cancer or melanoma. The dysplastic nevus syndrome carries with it a very high risk among persons with a family history of melanoma. Sus- picious moles should be surgically removed as soon as possible. The syndrome has also been found among persons without a family history of melanoma. The risk of melanoma is increased for these individuals. but is not so high as in those with a family history of melanoma. Such dysplastic moles should be watched carefully for any changes. though. and removed if changes take place. Thus. it appears that some melanomas result from inherited traits. Large congenital nevi. pres— ent at birth. appear to carry an increased risk of melanoma compared with moles that develop later in life. Another genetic link with melanoma is xe- roderma pigmentosum. a rare hereditary skin dis- ease that predisposes a person to all forms of skin cancer. Individuals with this condition lack an en- zyme that normally repairs cell DNA damaged by UV radiation. Hormonal factors may also play a part in melanoma. In one study. women who had used oral contraceptives for 5 years or more and those Risks for Major Cancers who had a first child after age 30 had a higher risk of melanoma than did a control group (Holly et al. 1983). But how these factors are related to melanoma is not clear. Despite its potential for serious disease, melanoma is highly curable when detected and re- moved early. In Australia, where melanoma inci- dence and “melanoma consciousness" are both high, at least 90 percent of patients live 5 years. These are some of the signs that may signal melanoma: I Changes in the size or color of a mole. or rapid darkening. I Ulceration or scaliness. changes in the shape or outline of a mole. or ifthe mole begins to flake. ooze. or bleed. I Changes in the way a mole feels—hard. lumpy. itchy. tender. I Changes in the skin around a mole. such as redness or swell— ing. I Erosion or scabs around a mole. I Development of a new. pigmented lesion or bulge in a normal skin area. Multiple Myeloma Multiple myeloma is a cancer of the plasma cells found in the marrow of the long and flat bones of the ribs, skull. legs, hips and spine. These cells normally produce immunoglobulins, or antibodies‘ that circulate in the blood and help to ward off dis- ease. ln multiple myeloma, the plasma cells pro- duce either too much of a single. specific immu- noglobulin or produce just part of an immunoglobulin. Symptoms of multiple myeloma include anemia, kidney failure, increased susceptibility to infection, and bone pain and fractures. The disease is also characterized by numerous bone lesions. These le— sions account for the name “multiple" myeloma and for the earlier belief that this was a bone can- cer. The most common form of multiple myeloma is secretory myeloma, in which the plasma cells se- crete a normal immunoglobulin at an abnormally high level. The immunoglobulins affected are IgG and IgA, and to some extent IgD and lgE. More than half of all patients with multiple myeloma have IgG secretory myeloma. Little is known about the causes of multiple my- l01 102 eloma. It is known that multiple myeloma occurs more often among aged, black men than among other U.S. population groups. The median age of onset for multiple myeloma in this country is 68 (Young et al, [981). It is the 13th most common type of cancer among U.S. blacks, but is ranked 19th among whites (Young et al, 1981). The incidence for black men is 9.6 cases per 100,000 and for black women, 6.7. For white men it is 4.3 and for white women, 3.0 (Young et al, 1981). Incidence varies worldwide. Blacks in the San Francisco Bay area have the highest recorded inci- dence; residents of Poona, India, have the lowest (Waterhouse et al, 1982). The incidence among all Asians is generally low. One theory to explain the predominance of mul— tiple myeloma in older individuals is that the im— mune system changes with age, and plasma cells are somehow altered, increasing their likelihood of becoming cancerous (Radl et al, 1975). The risk factors associated with race and gender have not been identified, and it is not known if these risks have a genetic or environmental basis. Blacks, for example, have higher levels of immunoglobulins than whites which suggests a possible inborn sus- ceptibility to multiple myeloma (Blattner, 1981). Men also have higher levels of immunoglobulins than do women, suggesting a possible hormonal link with the disease. Other factors reported to be associated with multiple myeloma include some occupational ex— posures, ionizing radiation, genetic susceptibility and immunosuppressive diseases. Some preliminary studies link certain occupa- tions with the development of multiple myeloma. Male compositors working for the U.S. Govern— ment Printing Office who were exposed to lead va- pors had a twofold increase in rate of death from multiple myeloma (Greene et al, 1979). An increase in multiple myeloma deaths was also seen among people who live in areas near plastics manufactur- ing industries (Mason, 1975). Plastics and oils have been shown to produce plasma cell cancers in mice. Farmers, wood workers, leather workers, workers exposed to arsenic, asbestos or lead, and employees of rubber and petrochemical products Risks for Major Cancers plants have shown an increase in multiple myeloma incidence (Blattner, 1982). One of the risk factors for multiple myeloma ap- pears to be radiation exposure. Radiologists with long-term, low-dose exposure to X-rays have shown a threefold increase in deaths from multiple myeloma (Matanoski et al, 1975). Atomic bomb survivors also show an increased incidence of mul- tiple myeloma (lchimaru et al, 1982). Radiation has an immunosuppressive effect that may favor the development of multiple myeloma and possibly other immune system disorders. Siblings of patients with multiple myeloma have an increased risk of developing the disease; this fa- milial relationship may be due to an inherited sus— ceptibility (McKusick, 1978). There is also some association between multiple myeloma and other illnesses that result in over- production of immunoglobulins. Such illnesses in- clude systemic lupus erythematosus, a chronic connective tissue disorder; scleroderma, a disease that causes hardening and thickening of connective tissue and skin; rheumatoid arthritis; and chronic dermatitis (Blattner, 1982). Non-Hodgkin’s Lymphoma Lymphomas are cancers that affect white blood cells ofthe immune system. They are charac- terized by the abnormal growth of lymphocytes, the infection-fighting cells in the lymph nodes, spleen, and thymus. The tonsils, stomach. small intestine, and skin may also be affected. Leuke- mias, in contrast, are cancers of circulating white blood cells that originate in the bone marrow. Lymphomas are usually classified as Hodgkin‘s disease, the most common form. or non—Hodgkin‘s lymphoma. There are also other rare forms of the disease such as mycosis fungoides‘ a primary skin lymphoma. Its appearance is often preceded by a history of chronic skin rash. Burkitt’s lymphoma. rare in most of the world, is the most common childhood cancer in central Africa, and is one of the fastest growing human cancers. The non-Hodgkin’s lymphomas are among the less common cancers in the United States. Inci- 103 l04 dence generally increases with age. Among white males, incidence is about 10.7 per 100.000. and 8.2 per l()0.()00 among white females. Incidence among black males is about 7.2. and 4.7 among black females (Young et al. I981). Five-year sur- vival for both white and black patients in this coun- try is about 43 percent (Ries. 1983). Incidence. sur- vival. and mortality statistics both here and abroad may be difficult to interpret because different sys- tems are used to classify these diseases. Non-Hodgkin‘s lymphoma may run in families but it is not known if this is due to heredity or a shared environmental factor. Some genetically de- termined immune system disorders can increase the risk of developing the disease (Purtilo. 1977). but most cases of familial non-Hodgkin's lympho- ma cannot be attributed to these genetic disorders. Patients with primary immune system disorders may also have a high risk of developing non- Hodgkin‘s lymphoma and certain other types of cancer (Gatti and Good. l97l). Kidney transplant patients. whose immune sys- tems are suppressed with medications. develop non-Hodgkin‘s lymphomas 40 to 100 times more often than expected (Fraumeni and Hoover. 1977: Kinlen et al. I979). The short latent period—onset is often within one year of the organ transplant— and the fact that transplant patients are more prone to certain viruses linked with cancer suggest that some viruses may play a role in non-Hodgkin‘s lymphoma (Greene. 1982). Epstein-Barr virus (EBV) and human T—cell leu— kemia-lymphoma virus (HTLV) have been linked with certain rare forms oflymphoma. EBV is ubiq- uitous—by adulthood almost everyone has been exposed to it and has developed antibodies against it. EBV has been linked to Burkitt‘s lymphoma in African children. HTLV is linked with certain types of adult T—cell leukemias and lymphomas that are found mostly in southern Japan. parts of the Caribbean and Africa, and the southeastern United States. Only a few of those exposed to this virus develop cancer. Researchers believe it can only be spread by prolonged intimate contact and is not highly contagious like other viral diseases. Occupational studies have been inconclusive. Workers exposed to the herbicides phenoxyacetic Risks for Major Cancers acid or chlorophenol have a higher than expected incidence of lymphomas (Hardell. 1979). Chemists also appear to have a high risk. but specific ex- posures have not yet been identified (Li et al. 1969: Olin and Ahlbom. 1980). Although non-Hodgkin's lymphomas account for only a small fraction of all cancers. they are re- garded as models for understanding the immune response and carcinogenesis (Greene. 1982). Oral Cavity and Pharynx Tobacco use—smoking. chewing and dipping—is the major risk factor for cancers of the mouth and throat. Depending on the amount and type of to- bacco used. there is a four to fifteenfold greater risk of developing mouth and throat cancers for to- bacco users over nonusers (Mahboubi et a1. 1982). Mouth and throat cancers are primarily squam- ous cell carcinomas. The most common sites are the tongue. lip, floor of the mouth. soft palate. ton- sils. salivary glands and back of the throat. More than 90 percent of all oral and pharyngeal cancers occur in individuals over age 45. and risk increases with age (Mahboubi et al. 1982). In certain parts of India, a large proportion of both men and women chew “pan.” a quid of betel leaves. nuts. tobacco and lime. Indians have the highest rates of mouth and throat cancers in the world and about 75 percent of these mouth and throat cancers can be attributed to tobacco chew- ing habits (Jayant et al. 1977). Indian women have a death rate from mouth and throat cancers 40 times greater than do American women (Mahboubi et al. 1982). Mouth and throat cancers occur at a rate of 16.8 cases per 100,000 population a year among US. white men. 60 among white women. 19.3 among black men. and 7.0 among black women (Young et a1. 1981). Even though men have higher rates of mouth and throat cancers. women are also susceptible to the cancer-causing effects of tobacco use. Women in the rural south who “dip snuff“ by holding a pinch of finely ground tobacco between the gum and cheek have a high risk of developing cancers of the 105 l 06 mouth and throat (Winn et al. I981). Snuff contains N—nitroso-nornicotine. a chemical known to cause cancer in mice (Hecht et al. 1980). Cigarette smoking. which is highly associated with lung cancer. also increases the risk for can- cers of the mouth and throat. Risk increases with increasing cigarette consumption. Heavy smokers (smoking more than one pack of cigarettes a day) are one-and-a-half times more likely than light smokers to develop mouth and throat cancers. and they have a sixfold increase in risk over non- smokers (Graham et al. 1977). Cancers of the mouth and throat most often de- velop at the site directly exposed to tobacco. Snuff—dippers develop cancers of the gum and the mucosal lining of the cheek: cigarette smokers de— velop more throat cancers; and pipe smokers de- velop more lip cancers (Smith. 1979). Alcohol drinking has been related to an increase in risk of mouth and throat cancers. Heavy drink- ers often smoke. however. and the effects of the two factors cannot always be separated. It is esti- mated that heavy drinkers who consume more than seven drinks a week have a doubled risk of mouth and throat cancers (Graham et al. 1977). Smoking and drinking have a synergistic effect; in most studies, the risk of mouth and throat cancers is greater for the combined factors of smoking and drinking than for the simple addition of the two (Mahboubi et al, 1982). Among those who drink at least 1.5 ounces of alcohol and smoke 40 or more cigarettes a day. there is a fifteenfold increase in risk of mouth and throat cancers (Rothman et al. 1972). Poor nutrition. possibly related to a lack of vitamins A and B. has also been linked to an in- crease in mouth and throat cancers. Some early studies have linked certain occupa- tions with the development of mouth and throat cancers. In an Australian study. bartenders. wait- ers and waitresses. presumed to be exposed to to- bacco smoke and alcohol on theirjobs. were found to have an increased risk of mouth and throat can- cers (McMichael et al. 1982). Printers (Lloyd et al. I977). leather workers (Decoutle. [979). paper manufacturers (Blot and Fraumeni. I977). elec- tronics workers (Winn et al. 1982). farmers. sailors, and outdoor workers are prone to lip cancer. There Risks for Major Cancers has also been some suggestion that metal. textile, and steel workers, and workers exposed to asbestos and polyvinyl chloride may have an increased risk of mouth and throat cancers (Mahboubi et al, 1982). Ill-fitting false teeth and bridges and sharp or broken teeth that can cause irritation or infection are also associated with an increased risk of mouth cancers. But the risk seems to be higher among those who also smoke and drink (Graham et al, 1977). Some slight evidence links the daily, long- term use of mouthwash among those who neither smoke or drink with mouth and throat cancers (Blot et al, 1983). Quitting cigarettes, pipes, and snuff would dras- tically reduce the incidence of cancers of the mouth and throat, but tobacco use—snuff dipping and tobacco chewing among teenagers, as well as cigarette smoking among women—appears to be increasing (Economic Research Service. I983: Horm and Asire, I982). Ovary In 1983, an estimated 18,000 new cases of ovarian cancer were diagnosed in this country (Silverberg and Lubera, 1983). Although ovarian cancer ranks second in incidence among gynecologic cancers, it causes more deaths—11,000 each year—than any other cancer of the female reproductive system (Young et al, 1981). The ovaries, two almond-sized glands containing egg cells, lie in the lower abdomen, one on each side of the uterus. By 5 months after fertilization. the ovaries of a developing female fetus already contain about 7 million egg cells, her entire life’s supply. The ovaries also secrete hormones that help regulate menstruation and pregnancy. Besides egg cells, the ovaries contain several other types of cells. Although cancer can affect any of these, 80 to 90 percent of ovarian cancers arise from the layer of epithelial cells that surround the ovary. Epithelial cancers develop in either ovary with about equal frequency, and develop in both ovaries at once about a third of the time. 1 ()7 108 The incidence of ovarian cancer in European and North American women has increased only slightly since the 19403. White women 40 to 50 years old living in highly industrialized countries develop the disease most often. In the United States. a woman has a [.3 percent chance of developing ovarian cancer by age 74 (Young et al. 198]). Childbearing is the most important known factor in preventing ovarian cancer. suggesting that hor- mones may play a role in its development. Women who have had children are half as likely to develop ovarian cancer as women who have not; several pregnancies confer even more protection. Use of birth control pills. which create a hormonal bal- ance similar to that found during pregnancy, may reduce the risk of ovarian cancer by 10 to 50 per— cent (Weiss. 1982). Breast cancer may also increase a woman‘s chance of developing ovarian cancer: women with breast cancer have twice the expected risk of de- veloping ovarian cancer. Women who already have ovarian cancer are three to four times more likely to develop breast cancer (Young et al. 1982). Studies of Japanese women in Hiroshima ex- posed to atomic bomb radiation during World War [1 revealed almost twice the expected number of ovarian cancer cases (Beebe et al. I977). But the X-ray doses used for diagnosis are not likely to in- crease a woman’s chances of developing the dis- ease (Weiss. I982). Exposure to asbestos has been linked to ovarian cancer risk in one study of women working in as- bestos-contaminated industrial areas (Newhouse et al. 1972). Particles of asbestos have been found in normal and cancerous ovaries. as have particles of talc. a mineral related to asbestos (Henderson et a]. 1972). Because mineral deposits of asbestos and talc are often found near each other. it is possible that talc may become slightly contaminated with asbestos during mining. Only two studies have ex— amined ovarian cancer risk associated with talc use in women. One study found an increased incidence in talc users (Cramer et al. I982): the other did not (Hartge et a]. [983). Talc has not been shown to cause cancer in laboratory animals (Hildick-Smith. 1976). Risks for Major Cancers Pancreas Cancer of the pancreas is a “silent“ disease. with- out symptoms. until it is advanced. Very little is known about what causes it or how to prevent it. An organ about six inches long located behind the stomach. the pancreas has two functions: it sends insulin into the bloodstream to control the amount of sugar in the blood, and sends pancreatic juice into the intestine to help digest food. Small tubes or ducts in the organ transport the pancreatic juice and if cancer develops it is usually in these duct cells. From l95| to 1978. the death rates for pancreatic cancer in the United States rose almost 30 percent to about 11 deaths per 100,000 men and about 7 per 100,000 women (American Cancer Society. 1982). The incidence of pancreatic cancer among US. blacks is about 1.5 times higher than for whites (Young et al. 198]). Hawaiians and American Indi- ans are also at a higher risk (Young et a1, I981). Seventh—Day Adventists and Mormons have a lower-than-average death rate (Phillips et al. I980; Lyon et al. 198]). and Jewish men have a higher- than-average death rate (Greenwald et al. 1975). Worldwide. the United States and Northern Euro- pean countries. including Great Britain. have a high incidence of pancreatic cancer. Polynesians have an extremely high incidence. The disease is usually fatal; only 4 percent of pa- tients live more than 3 years after diagnosis (Ries et al. I983). The very few patients whose cancers occur in the insulin—producing cells—not the duct cells—tend to live longer; about 30 percent of these patients live more than 3 years after diag- nosis (Axtell et al. 1976). After 30. the incidence of this cancer increases in both men and women in every population stud- ied (Mack. I982). An excess risk has been estab— lished among cigarette smokers. although the mag- nitude of the risk is not so great as with lung cancer. Diabetes mellitus has been linked in some studies with pancreatic cancer. but it is not known if cancer causes diabetes-like changes, or if di— abetes makes individuals prone to cancer. No clear association has been found between diet and pan- creatic cancer. Coffee drinking has been associated with the disease in one study but that study has not been confirmed (MacMahon et a1. 1981). 109 Risks for Major Cancers Research has focused on ways to diagnose pan- creatic cancer before it is advanced enough to cause symptoms. Ultrasound and CAT scans are being tried. but to date only a biopsy yields a cer- tain diagnosis. Surgery. radiation therapy. and anti- cancer drugs are used to treat pancreatic cancer. but so far have had little influence on outcome. In 1975. the National Cancer Institute established the National Pancreatic Cancer Project to stimulate re- search on the causes. diagnosis. and treatment of the disease. Prostate “0 Cancer of the prostate is one of the most common cancers among United States men. lts incidence increases with age and it is chiefly a disease of men over age 65. The prostate gland. located at the base of the penis, surrounding the urethra. produces seminal fluid. Two conditions that commonly affect it are enlargement and cancer. but the two do not seem to be related to each other Black men in the United States have the highest incidence in the world of cancer of the prostate (Greenwald I982). Between [975 and 1977. the In- cidence among black men in Atlanta was about I33 per 100.000. compared with about 74 per 100.000 for white men in the same city (Young et al. 198]). The high incidence of this cancer among blacks has occurred only 1n the last few decades suggesting that social factors rather than genetic factors are responsible (Ernster et al I978). Cancer of the prostate is also common in north— west Europe. Incidence is lower in the Near East and in parts of Africa and South America. The lowest incidence occurs in Japan. Studies of migrating populations have suggested that environmental factors such as diet and life- style play an important role 1n the risk of develop- ing cancer of the prostate. Prostatic cancer inci— dence and fat intake are higher among Japanese in Hawaii than in Japan, for example (Kato et al, 1973: Waterhouse et al. 1976). Mormons who do not use tobacco. alcohol, cof- fee. or tea for religious reasons. but whose fat con- Risks for Major Cancers sumption is similar to that of other white males in the United States. have about the same risk of can- cer of the prostate as other white men. It is the most common type of cancer among Mormons (Enstrom. I978). Prostate growth and function depend on the hor- mone testosterone. formed in the testicles. It is possible that diet affects the production of sex hor- mones. and that this may then affect the risk for cancer of the prostate (Graham et al. 1983). Cancer of the prostate. as well as cancers of the colon. rectum. and female breast. may be associ- ated wtih dietary fat intake (Wynder et al. 1971: Carroll and Khor. 1975; Berg. 1975). Cancer of the prostate has been linked with the consumption of animal fat and protein among several ethnic groups in Hawaii (Kolonel et al. 198l). Workplace exposures to cadmium during weld- ing. electroplating. and the production of alkaline batteries may increase the risk of cancer of the prostate. Dietary exposures to cadmium. from oysters for example, do not seem to increase risk (Kolonel and Winkelstein, 1977). Workers in the rubber industry may also be at increased risk (Tyroler et al. 1976). Thus. the causes of cancer of the prostate are unclear. so there are no known methods of preven- tion. The possible role of diet and of workplace carcinogens are now being studied. Skin Cancer (Nonmelanoma) Nonmelanoma skin cancer is the most common cancer among whites in the United States. Since most nonmelanoma skin cancer patients are treat- ed in doctors” offices. population—based estimates of skin cancer incidence are fairly difficult to ob— tain. Estimates are, though. that more than 400.000 new cases of nonmelanoma skin cancer occur in the United States each year, and that this number is rising (Fears and Scotto, 1982). Although the death rate from nonmelanoma skin cancer is about 1 percent. as many persons died from this cancer in the 19508 and 1960s. for example. as from the rarer, but more lethal. skin melanoma (Mason et al. 1975). III The incidence of nonmelanoma skin cancer var- ies directly with exposure to ultraviolet (UV) light from the sun and indirectly with the degree of skin pigmentation. Thus. nonmelanoma skin cancer is most common among fair-skinned whites who live in sunny locales. The highest rates in the past have been recorded among Caucasians in South Africa and Australia. Even lreland. despite its rain and mist. has had a high incidence because ofthe sus- ceptibility of persons of Celtic ancestry (Urbach. 1971). Nonmelanoma skin cancer occurs less often in Orientals, and least often among blacks. In the United States. for example. a l977—78 National Cancer Institute survey (Scotto et al. 1981) showed that the age-adjusted incidence was only 3.4 per 100.000 among blacks. compared with 232.6 among whites. Most nonmelanoma skin cancers are of two types. squamous cell carcinoma and basal cell car— cinema. The basal cell type is more common. but the squamous cell type is more invasive. and may account for about three-fourths of all deaths from nonmelanoma skin cancer (Dunn et al. 1965). Based on the 1977—78 US. skin cancer survey. basal cell cancer occurs about I '/2 to 2 times more often in white men than in white women, and squamous cell cancer occurs two to three times more often in men. Both types occur most often on the face. head and neck. Women have higher rates than men for both types of cancers on the legs. in line with their greater sun exposure. while men have more squamous cell carcinoma of the lip. in line with their risks from tobacco and outdoor work (Lindqvist, 1979). When skin cancer incidence is plotted for US. cities according to annual UV—B measurements. the direct relationship is most clearly seen with squamous cell cancer and increasing radiation (Scotto and Fraumeni. I982). Basal cell cancer in- cidence also reflects the increase in radiation. Melanoma incidence follows it least sharply. This is consistent with the evidence that factors other than sunlight also contribute to the development of melanoma (Greene and Fraumeni. [979). Thus the chief risk factor for nonmelanoma skin cancer is exposure to nonionizing UV radiation. most probably the UV-B portion of sunlight radia- tion. and evidence suggests that the risk increases with the annual dose (Fears and Scotto. I983). Pos- sible depletion of the protective ozone layer by the fluorocarbons used in aerosol propellants. re- frigerators. and air conditioners is a matter of some concern. There are other risk factors for nonmelanoma skin cancers. They were, for example. the first type of cancer related to ionizing radiation. with reports as early as 1902 among radiation workers. Other studies have shown an excess risk associated with radiotherapy for a number of diseases. Excess risks have also been noted among radiologists and uranium miners. Radiologists were at one time ex— posed to X-rays from their own equipment. and uranium miners were exposed to radon daughters. the radioactive products of uranium ores. A number of chemicals induce skin cancers. par— ticularly squamous cell carcinomas in animals. and epidemiologic studies substantiate their risk in hu— mans. Polycyclic aromatic hydrocarbons induce cancers in animals and are found in coal tars. pitch. asphalt. soot. creosotes. and lubricating and cutting oils. Skin and other forms of cancer have been found in various worker groups exposed to these substances. A recent study has shown an ex- cess risk of skin cancer among psoriasis patients treated with crude tar ointments (Stern. I980). Psoralens. which render skin more sensitive to light. have also been used in combination with ul- traviolet light A (PUVA) for psoriasis patients. and an excess risk of skin cancer has been seen among these individuals also (Stern et al. I979). This study has increased concern about the possible hazards of other photosensitizers found in tanning aids, cosmetics. and medicines. Squamous cell skin cancer has also been found as a complication of tropical ulcers. burns. scars. and chronic infections and wounds (Malik et al. 1974). This complication has been seen chiefly among dark-skinned populations in Africa and Asia. but recent studies of black Americans have indicated that burn scars or chronic infections may predispose them to skin cancer. Actinic ker- atoses—brownish. hardened areas on skin exposed to excess sunlight—are considered to be precursor ll3 Risks for Major Cancers lesions for squamous cell skin cancer. Individuals with several rare hereditary diseases—including multiple basal cell carcinoma syndrome. xeroder— ma pigmentosum. and albinism—are also at height- ened risk of skin cancers. Avoiding overexposure to sunlight is the chief way to prevent nonmelanoma skin cancer. It is also important to avoid unnecessary X—rays and ultra- violet light exposure from artificial sources like sunlamps and tanning booths. Stomach 114 In the 19305. stomach cancer was the leading cause of cancer death among US. men. and the third leading cause among US. women after cancers of the uterus and breast. Since then. there has been a dramatic decrease in stomach cancer death rates in both sexes. The death rate decreased about 60 per- cent between 1951 and 1978. and today. U.S. death rates for stomach cancer are among the lowest in the world (American Cancer Society. I982). Despite the decrease. stomach cancer is still a major problem. About 25.000 new cases of stom- ach cancer were expected in the US. in 1983 (American Cancer Society. I982). but only about 13 percent ofthese patients will live for 5 years after diagnosis (Ries et al. 1983). Stomach cancer has had one of the poorer 5-year survival rates of any type of cancer in the US. (Ries. et al. 1983). One reason is that it may not be detected until an advanced stage. when it has spread to other parts of the body. Japan. Chile. Costa Rica. Colombia. Singapore. and Iceland are among the countries with high death rates from stomach cancer. In Japan. stom- ach cancer causes more deaths than all other types of cancers combined. and it is five times more common than in the US. When Japanese move to the United States. though. their stomach cancer rates decrease over successive generations. Migrant studies and the marked decrease in US. death rates suggest that environmental factors play a dominant role. The widespread use of refrigera- tion since the I930s may be partly responsible for the reduced rates. Refrigeration reduced the need for some other methods of food preservation and gave Americans access to fresh fruits and vegeta- bles year—round. International surveys have shown an association between stomach cancer and large amounts of pickled. salted or smoked foods in the diet (Nomura, 1982). Other studies suggest that stomach cancer patients eat fewer servings of foods rich in vitamins C and A than do control subjects (Correa et al. 1982). Persons with diseases that af— fect the stomach lining. such as pernicious anemia and atrophic gastritis. have a higher risk of devel— oping stomach cancer than the general population (Correa. I982). but gastric ulcers have not been consistently associated with stomach cancer. Nitrosamines are potent carcinogens that can form in the stomach. When nitrates, a family of chemicals found in some water supplies and in some green vegetables. cured meats. and cheeses. combine with bacteria in the mouth. different com- pounds known as nitrites may form. The nitrites. in turn, combine with components of some foods. drugs. and other substances to form carcinogenic nitrosamines. Vitamin C appears to be a natural defense against nitrosamines by preventing their formation in the body. Studies in several countries have shown that eating fresh fruits and vegetables containing vitamin C reduces the risk of stomach cancer (Nomura. I982). There may be other environmental and lifestyle factors involved. Studies in Connecticut. Hawaii. Norway. Iceland. and Japan have consistently shown that low socioeconomic status is associated with stomach cancer (Nomura. I982). An associa- tion with cigarette smoking has also been sug- gested (Haenszel et al. 1972; Hammond. 1966). Ev— idence that radiation may increase risk comes from studies of atomic bomb survivors in Japan and of patients treated with X-rays for a spinal disorder. Evidence for a genetic component comes from studies suggesting that blood type A. an inherited trait, may be a marker for increased risk (Nomura. 1982). Other research shows that relatives of per- sons who had stomach cancer have a risk two to three times higher than the general population (Nomura. 1982). Families tend to share the same environment. though. so it is difficult to distinguish between genetic and environmental influences. ll5 Risks for Major Cancers Reducing the amounts of pickled. salted. or smoked foods. and of nitrates consumed. may re- duce the risk of stomach cancer. A diet rich in fresh fruits and vegetables may also help reduce risk. Testis H6 American men have only a 0.3 percent chance of developing testicular cancer in their lifetimes (Young et al. I981). But among young white men aged 20 to 34. testicular cancer is the most com- mon form of cancer. accounting for 22 percent of all cancers in this age group. It is the second most common cancer among men aged 35 to 39. and the third most common among young men aged 15 to 19 (Schottenfeld and Warshauer. 1982). The testes are small. oval glands that produce sperm and testosterone. the male hormone. The testes form in the abdominal cavity early in fetal development and usually descend to the scrotal sac before birth. Almost all testicular cancers are germ cell can- cers; the most common of these are a particular type called seminomas (Schottenfeld and War- shauer. I982). The germ cells are the sperm-form- ing cells of the testes. Men in rural Vaud. Switzerland. have the highest incidence of testicular cancer in the world with 10.5 cases per 100,000 (standardized to world pop- ulation) a year; Cuban men have the lowest inci- dence. near zero (Waterhouse. 1982). Among U.S. white men. the incidence is 3.6 cases per l00.()()0 a year. This is over four times the rate (0.8) for U.S. blacks (Young et al. [981). The incidence of testicular cancer among His- panics. native Americans and Asians lies between those of white and black men. The death rate from testicular cancer among U.S. white men aged 20 to 29 declined about 40 percent between 1973 and 1978 even though the in— cidence for this high—risk group increased slightly during this time period (Li et al. 1982). The reasons for the upward trend in incidence are not clear, but the decline in deaths is due mainly to the increase in survival brought about by advances in testicular cancer treatment. The outlook for men with semi— nomas found early is very good. Little is understood about the causes of testicu- lar cancer. but men aged 20 to 34 years are at the greatest risk. Possible risk factors are congenital abnormalities. hormonal drugs. and trauma. Testicular and genital abnormalities have both been associated with testicular cancer. Cryp- torchidism. failure of the testes to descend into the scrotal sac. is thought to account for one in 10 cases of testicular cancer (Shottenfeld and War- shauer. 1982). Studies have also linked inguinal her- nia in children with adult onset of testicular cancer (Morrison. 1976). There is some evidence that these hernias are due to incomplete descent of the testes. so they should not be considered separately from cryptorchidism (Coldman et al. 1982). Other conditions associated with testicular cancer in- clude some rare genetic abnormalities (like Klinefelter‘s syndrome. hermaphroditism and Turner's syndrome): gonadal aplasia. or failure of the gonads to develop; hypospadias. a condition in which the urethra opens on the underside of the penis: and mixed gonadal dysgenesis. a condition in which there is one developed testis plus non- functional female genitalia (Schottenfeld and War- shauer. 1982). Both the hormone estrogen and the synthetic es— trogen diethylstilbestrol (DES) injected into preg- nant mice can cause testicular abnormalities in male offspring (Henderson et al. 1979) DES ex- posure before birth has been linked to vaginal can- cer in daughters. and to testicular abnormalities in sons of women who took it to prevent miscarriages (DES Task Force. 1978). If DES or estrogens in- crease the risk for testicular cancer in men is not yet clear. Other factors that may be related to increased risk of testicular cancer are maternal bleeding dur- ing pregnancy and history of stillbirths (Swerdlow et al. 1982). A recent study suggested that teenage participa- tion in sports like bicycling and horseback riding may be associated with testicular cancer. but more study is needed before it can be concluded that the risk observed is a real one. The study also did not yield evidence that might explain the possible asso- ll7 Risks for Major Cancers ciation. Scientists have speculated from time to time that trauma might somehow increase a man‘s risk of testicular cancer, but no studies to date have shown a definite association. There is some contradictory evidence concern- ing testicular cancer and socioeconomic status. Some studies have found that those with high in- come and high education are two-and-a-half times more likely to develop testicular cancer than those with less income and education (Schottenfeld and Warshauer. 1982). There is also some evidence that men from rural areas have higher rates of testicular cancer than city dwellers (Graham et al. I977). An association between exposure to viral disease like mumps orchitis and adult onset of testicular cancer has not been proved although an early case report in the 1940s suggested such a risk (Gilbert. I944). Urinary Tract ll8 Urinary tract cancers account for 9 percent of the new cancer cases diagnosed each year in men and 4 percent of those in women (Silverberg and Lubera, I983). The two most common urinary tract cancers are bladder cancer and kidney cancer. The estimated 38,500 cases of bladder cancer that Americans developed in 1983 make it the 6th most common cancer in this country. Kidney cancer. es- timated to occur in 18,200 Americans in [983. ranks 11th in cancer incidence. Besides the two kidneys. the urinary tract in— cludes the ureters that carry urine from the kid- neys to the bladder. the bladder, and the urethra. a tube that carries urine from the bladder. This sys- tem of organs eliminates liquid waste products and helps maintain stable chemical conditions in the fluid that surrounds body cells. Bladder In the United States, bladder cancer is chiefly a disease of white men over age 65. The incidence of 27 cases per 100.000 population among white men is twice the incidence among nonwhite men. There is little racial difference in incidence among wom- en, who develop bladder cancer less than a third as often as men do (Young et al, 1981). From the late 19405 to the early 19705, the inci- dence of bladder cancer declined 21 percent for white women and 38 percent for nonwhite women. At the same time, it increased 24 percent for white men and doubled for nonwhite men (Devesa and Silverman, 1978). Around the world, bladder cancer occurs most often in the United States and Europe and least often in Asia (Waterhouse et al. 1982). The most important known risk factor for blad- der cancer is cigarette smoking. Cigarette smokers develop bladder cancer two to three times more often than nonsmokers, and areas in the United States where cigarette sales are high also have high death rates from bladder cancer (Morrison and Cole, 1982). Smoking is estimated to be responsi- ble for about 40 percent of the bladder cancers among men and 29 percent among women (Cole et a1, 1971). As early as 1895, workers in the dyestuffs indus- try showed a high risk of bladder cancer that was later associated with exposure to aromatic amines, a class of compounds used to make dyes (Rehn, 1895). Two of these chemicals, benzidine and 2- naphthylamine, are now known to be potent blad- der carcinogens (Case, 1954). Workers in the rub- ber and leather industries also have an increased risk of developing bladder cancer. Occupations in which workers are suspected of having an elevated bladder cancer risk include painter, chemical work- er, printer, metal worker, hairdresser, textile work- er, machinist (Morrison and Cole. 1982) and truck driver (Silverman et a1, 1983). The possible risk of bladder cancer associated with widely used artificial sweeteners received much attention when the Food and Drug Adminis- tration removed cyclamates from the market in 1969. It was later reported that the sweetener sac- charin caused bladder cancer in male laboratory rats when the animals were exposed to the chem- icals before birth (Arnold et a1, 1977). But recent epidemiological studies show that, overall, people who use artificial sweeteners do not appear to have a higher incidence of bladder cancer than non-users (Morrison and Buring, 1980; Hoover et al, 1980). 119 I20 Although early studies showed a possible link between bladder cancer and coffee drinking, recent studies based on large numbers of individuals found little or no increase in bladder cancer inci- dence among coffee drinkers compared with those who do not drink coffee (Morrison et al, 1982: Hartge et al, 1983). Other factors that may contribute to the develop- ment of bladder cancer are bladder infection with the parasitic fluke Schistisoma haematohium, treatment with the anticancer drugs chlor- naphazine or cyclophosphamide and long—term use of pain killers containing the drug phenacetin (Mor- rison and Cole, 1982). Kidney About 85 percent of the kidney cancers diagnosed in this country are renal cell cancers. Cancer of the renal pelvis, the inner part ofthe kidney connected to the ureter, accounts for most ofthe remaining 15 percent. The incidence of renal cell cancer among white men. 9.4 cases per 100,000 population, is nearly the same as for black men. 8.7 per 100,000. Renal cell cancer occurs twice as often in men as it does in women. and develops most often in both sexes in persons over age 60 (Young et al, 1981). Worldwide, the incidence of kidney cancer is high in North America and low in Asia (Water- house et al, 1982). As with bladder cancer, cigarette smoking is the most important known risk factor for kidney can- cer. Smokers are twice as likely as nonsmokers to develop kidney cancer (Wynder et al, I974; McLaughlin et a], 1984). One estimate is that 30 percent of kidney cancers in men and 24 percent in women are caused by cigarette smoking (McLaughlin et al, 1984). Among women, obesity, or factors associated with it, appears to be a risk factor for kidney can- cer (Wynder et al, 1974; McLaughlin et al, in press). Because fatty tissue can convert other hor- mones into estrogen, obese women may have high levels of this hormone (MacDonald and Siiteri. 1974; Schindler et al, 1972). The higher-than-ex- pected incidence of kidney cancer among obese Risks for Major Cancers women could be due to excess estrogen levels. There have been few studies of occupational risk factors for kidney cancer. Workers exposed to in- sulation fibers such as asbestos and workers in the petroleum industry show an elevated incidence of kidney cancer (Selikoff et al. 1979; Thomas et al. 1982). Although cancer of the renal pelvis is a fairly rare form of kidney cancer. one study reported that approximately 82 percent of the cases among men and 61 percent among women could be prevented if people stopped smoking (McLaughlin et al. 1983). Other factors that may increase the risk of cancer of the renal pelvis are the long-term use of pain relievers containing phenacetin or acetaminophen (McLaughlin et al. 1983: Morrison and Cole. 1982). Uterine Cervix Cancer of the uterine cervix. or cervical cancer. has been studied extensively. but no one cause has yet been found for it. A number of factors appear to contribute to risk. The incidence of invasive cancer of the uterine cervix and mortality from it have been declining steadily in this country for the past three decades. The incidence is still almost 2 '/2 times higher in US. black women than in whites. though. and the mortality is still almost three times higher in blacks. despite the declines in both groups. The uterine cervix is the small cylindrical neck that leads from the uterus. or womb. into the vagi- na. A knob of the cervix protrudes into the vagina and can be seen during physical examination. Cell samples are taken from this part of the cervix for the Pap smear test used to detect changes in cell structure that may lead to cancer. Researchers think that these cell changes. or dysplasias, may precede carcinoma in .w'tu, or CIS. which may then develop into invasive cancer of the cervix. This has not been proved. though. An argument for this relationship is that invasive cervical cancer occurs most often among women over age 50. and carcinoma in situ occurs most often among women 25 to 34 years old (Young et 121 [22 al, I981). Also, the incidence ofinvasive cervical cancer has been decreasing while CIS incidence has risen. This might mean that Pap smears are de- tecting more carcinomas in sin: and that treating these has prevented the development of invasive cervical cancer. Part of the decrease seen in cer- vical cancer incidence, though, might be due to the large number of hysterectomies—removal of the uterus and cervix—performed in the older age groups. In carcinoma in situ, an outer layer of normal cells has been replaced by cancer cells. It is about 95 percent treatable and curable. In invasive can- cer of the cervix, the cancer cells have invaded the underlying tissue ofthe cervix. Whether or not CIS and invasive cervical cancer are part of a continuum representing stages, or de— grees of cancer, the risk factors appear to be sim- ilar. The two major risks are multiple sex partners and early age at first intercourse. Early first inter— course is thought to be risky because the tissue of the cervix changes during puberty and may thus be more sensitive or vulnerable in young women. One study separated age at first intercourse from number of partners and found number of partners was more important (Harris et al. 1980). Frequency of intercourse with one partner does not appear to influence risk (Rotkin, I967; Wynder et aI, I954). There is also some suspicion that cervical cancer may be associated with venereal disease. This sus- picion has been strengthened by the finding that in~ creased levels of antibodies to genital herpes virus have been found in some women with cervical can- cer. indicating that they were exposed to herpes. Papilloma virus and Chlamydia, which cause geni- tourinary warts and infections in both men and women. are also being looked at in this connec— tion. Type of contraception is related to cervical can cer incidence; more cervical cancer is seen among women who use oral contraceptives. less among those who use barrier methods. This may be be- cause barrier methods protect against venereal in- fection, or because women who use oral con— traceptives may be more sexually active. The circumcision status of the man has been Risks for Major Cancers found not to be a risk factor. Smoking is a risk factor, although how it might be related biologically is unclear. The effects of nutrition on risk are not well es- tablished. Intake of foods high in retinol and car- otene has been found to exert a protective effect against some squamous cell tumors (the type that accounts for most cervical cancer), and two other studies have suggested that vitamin C and folicin— one of the B complex vitamins—are associated with decreased risk of cervical dysplasia and C13. Overall, the evidence suggests that cervical can- cer is caused by a number of agents, not just one. The National Cancer Institute (Brinton et al) is now doing a case control study of 2,000 women— 500 with C18, 500 with cervical cancer, and 1.000 controls—in Birmingham, Miami, Philadelphia, Chicago, and Denver that may help to shed more light on the causes of cervical cancer. Uterine Corpus (Endometrium) Cancer of the uterine corpus, or endometrial can- cer, is the third most common cancer among U.S. women and accounts for about 9 percent of all can~ cers in American women (Young et al, 1981). The uterus is a pear-shaped organ that lies in the abdomen between the bladder and the rectum. It consists of the cervix. the opening of the uterus into the vagina. and the corpus, sometimes called the body or womb. The corpus is composed of two layers of tissue. The spongy inside layer, the endo- metrium, proliferates between menses and is shed during menstruation if fertilization has not oc— curred. The outside layer, the myometrium, is a muscle capable of expanding during pregnancy to accommodate a growing fetus. Female sex hor- mones, including estrogen, prepare the uterus for pregnancy. Because most cancers of this site origi- nate in the endometrium, cancer of the uterine cor- pus is usually referred to as endometrial cancer (Young et al, 1981). After a large increase in the incidence of endo- metrial cancer in the 1970s, both the incidence and the death rate are now dropping (Austin et al, I982). ln U.S. white women. the annual incidence 123 l24 is 29.9 cases per 100,000. one of the highest in the world. and in US. black women. it is l4.6 (Young et al. 1981). Most women diagnosed with endo- metrial cancer are around age 60: the median age for white women is 61 and for black women. 64 (Young et al. 198] ). Some of the risk factors for endometrial cancer are the same as those for breast cancer: women at increased risk of developing breast cancer are also at increased risk of endometrial cancer. These risk factors include obesity, few or no children. early menarche. late age at menopause. and high socio- economic status (de Waard. 1982). Most ofthe risk factors for endometrial cancer may be related to hormonal imbalances. especially excess estrogen production (Elwood et al. 1977). Obesity has long been recognized as a risk factor for endometrial cancer. Obese women are twice as likely to develop endometrial cancer as women of normal weight. and this risk increases with increas- ing weight (Elwood et al. I977). There is also some evidence that obese women who are tall (5 feet 7 inches and over) are at even greater risk (Elwood et al. I977). Both diabetes and high blood pressure have also been associated with increased risk of endometrial cancer. but as both illnesses are relat— ed to obesity. it is not clear if these are separate risk factors (de Waard. 1982). Multiple births decrease risk. Women who have four or more children are one-third as likely to de- velop endometrial cancer as women who have no children (Elwood et al. I977). Women who have never had children. particulary women with a his- tory of infertility. are at greatest risk. Studies of women with Stein-Leventhal syndrome. a rare ill- ness characterized by multiple ovarian cysts. ex- cessive estrogen production. and infertility. are helping scientists define possible reasons for the observed reproductive associations (de Waard. 1982) Women of high socioeconomic status have an in— creased risk of developing endometrial cancer: diet and lifestyle may be contributing factors (de Waard. I982). Estrogen replacement therapy has been linked to endometrial cancer. Postmenopausal women who use estrogens are estimated to have a six- to seven- fold increase in risk (Antunes et al. 1979). The risk increases with increasing duration of use and dos- age of replacement estrogens (Antunes et al. I979). The use of estrogens for treatment of menopausal symptoms increased until about 1975. when their use decreased after reports associating menopausal estrogen with endometrial cancer (Jick et al. 1979). There was a subsequent decrease in endometrial cancer incidence (Austin et a1, 1982). Obesity has been linked to estrogen imbalance and there is a strong association between high lev- els of estrogen and development of endometrial cancer. Obese women. particulary after meno- pause. have higher levels of estrogens in their blood than women of normal weight. It is thought that this estrogen is produced in the excess fatty tissue (de Waard, 1982). Recent evidence shows that use of birth control pills may decrease the risk of developing endo- metrial cancer (CDC. I983). Women who use com- bination pills containing both estrogen and proges- terone in each pill for at least one year have only half the risk of endometrial cancer as women who use other types of birth control pills or none. “Sequential" pills. for instance, which contain a series of 16 estrogen and five progesterone pills, appear to double the risk of endometrial cancer (CDC. I983). The longer a woman takes the com— bination pill, the more this protection increases (Hulka et al. 1982). Childless women are protected even further by this pill and are two-and—a-half times less likely to develop endometrial cancer (Hulka et al, 1982). Glossary Age—adjusted. Cancer risk increases signifi- cantly with age. As the proportion of older individuals increases in any population. so. too. does the overall number of cancer cases. Data are age-adjusted mathe— matically to compare statistics over time in a given population. or to compare statistics among countries with different age propor— trons. Benign. Not life-threatening. Biopsy. Removal and examination. usually microscopic. of tissue or other material from the living body for purposes of diag- nosis. (‘areintme/L Any cancer-causing substance or agent. ("(n'areintmen. Any agent that increases or augments the effect of a carcinogen. (‘areinoma. A cancer that arises in the epi— thelial cells that cover external and internal body surfaces. (‘ase-eontro/ studies. Those that compare data on individuals with an illness (cases) with apparently similar healthy individuals (controls). matched by age. sex. and other factors. in an effort to define risk factors for an illness. ('o/iort. Any group of individuals selected for study. l26 Epidemiology. A science dealing with rela— tionships of various factors that determine frequency. distribution and possible causes of a disease in a human community, III(‘i(/('Il('('. New cases of a specific disease occurring during a certain period. usually expressed as new cases per 100.000 popula— tion per year. Initiator: A substance or agent that can start the process of carcinogenesis. Inyerxe a.\'.\'o('iution. Opposite in order or effect to that which is under consideration. Lateney or latent period. The incubation period of a disease. from exposure to disease development. Lesion. Any morbid change in the struc— ture of organs or parts. Leukemia. Cancer of the blood-forming organs. Lymphoma. A cancer that arises in lymph tissue. Malignant. Tending or threatening to produce death: opposed to benign. Metastasis. Spread of cancer cells from a primary tumor to sites elsewhere in the body. Morbidity. Illness. Mortality. Number of deaths occurring dur- ing a certain period. usually expressed as number of deaths per 100.000 population per year. Mutugen. A chemical or physical agent that induces permanent. transmissible genetic change. NCHS. National Center for Health Statis— tics. A Federal agency. part ofthe U.S. Public Health Service. that collects and analyzes data on health and disease. Neoplasm. Any new tissue growth. Populution—buxecl. Disease incidence and mortality data. for example. based on the exact count of a population in a given geographic location. Promoter. A substance or agent that com— pletes the carcinogenic process after initiation. Precursor. Forerunner: sign or indication that precedes. Predi.\'po.\‘e. To dispose or incline before- hand; to give a tendency to. Prognosis. Forecast of the course of a disease: outlook for it. Proweetive. Describing a study that begins with a present status of individuals in a par- ticular group and periodically reviews their status as time goes by. Relative risk. A measure ofthe risk ofdis— ease in an exposed group compared with the risk in an unexposed group. A relative risk of l.() means risks in the two groups are the same. lf. for example. risk is double for an exposed population. the relative risk is 2. Retrospeetive. Describing a study that tries to ascertain data based on past events. Risk/benefit. The relation between the risks and the benefits of a given treatment or procedure. Sureomu. A cancer that arises in connec— tive and skeletal tissue. i.e. muscle. bone. SEER. The National Cancer lnstitute‘s Sur— veillance. Epidemiology. and End Results program. begun in I973. lt monitors annual cancer incidence and survival in the United States. The original registries were made up of a It) percent nonrandom sample of the population. representing diverse popu- lation subgroups. ln I983. SEER was ex- panded to represent l2 percent of the pop- ulation in six states. four metropolitan areas. and Puerto Rico. .S'urvival. Usually expressed as a ratio: those who survive a disease per number of persons diagnosed with the disease in a given time period. .S'.\'nergi.\'n1. Cooperative effects of two agents giving a total effect greater than the sum of the two effects taken independently. USBC. U.S. Bureau ofthe Census. Census population counts are used for incidence and mortality rates and for the U.S. stan— dard population. References Adamson RH. Sieber SM: Antineoplastic agents as po- tential carcinogens, In Origins of Human Cancer lHiatt HH. Watson JD. Winsten JA. eds). Cold Spring Harbor Conferences on Cell Proliferation. pp 429—443. I977. Albores—Saavedra J. Alcantra-Vazquez A. Cruz-()rtiz H et al: The precursor lesions of invasive gallbladder carcinoma. Hyperplasia. atypical hyperplasia and carcinoma In xim. Cancer 45:919—937. I980. American Cancer Society: Cancer Facts and Figures. I983. New York: ACS. 1982. American Cancer Society: Cancer Facts and Figures, I984. New York: ACS. I983. Antunes CMF. Stolley PD. Rosenshein NB et al: Endo- mctrial cancer and estrogen use: Report of a large case-control study. N Engl J Med 300(II:9—I3. I979. Archer VIE. Gillam JD. Wagoner JK: Respiratory dis- ease mortality among uranium miners. Ann NY Acad Sci 271280—293. I976. Armstrong BK. Doll R: Environmental factors and can- cer incidence and mortality in different countries. with special reference to dietary practices. Int J Can- cer I526l7—63I. I975. Arnold DL. Moodie CA. Grice HC et al: Long—term toxicity of orthotoluene sulfonamide and sodium sac— charin in the rat. An interim report. Ottawa. Canada: National Health and Welfare Ministry. I977, Austin DI“. Roe KM: The decreasing incidence ofendo- metrial cancer: Public health implications, AmJ of Public Health 7Z(I):65—68. I982. Axtell LM. Asire AJ. Myers MH. eds: Cancer Patient Survival: Report No. 5. DHEW Pub] No. (NIH) 77-992. Washington. D.C.: U.S. Govt Print ()ff. I976. Beasley RP: Hepatitis B virus as the etiologic agent in hepatocellular carcinoma. epidemiologic considera- tions. Hepatology 1218—268. I982. Beasley RP. Lin CC: Hepatoma risk among HBsAg Carriers. AmJ Epidemiol 1081247. I978, Beebe GW. Kato H. Land CE: Mortality experience of atomic bomb survivors. 1950—74. Life Span Study. Report 8. Radiation Effects Research Foundation 'I‘RI—77. I977. Beebe GW. Kato H. Land CE: Studies ofthe mortality of A-bomb survivors. 6. Mortality and radiation dose. 1950—1974. Radiat Res 75:138—201. I978. BEIR. Advisory Committee on the Biological Effects of Ionizing Radiations (The BEIR Report). National Academy of Sciences-National Research Council. The Effects on Populations of Exposure to Low Lev- els of Ionizing Radiation. Washington. D.C.: U.S. Govt Print ()II‘. I972. I980. Berg JW: Can nutrition explain the pattern of interna- tional epidemiology of hormone-dependent cancers'.‘ Cancer Res 35:3345—3350. I975. I28 Berg JW: The incidence of multiple primary cancers. 1. Development of further cancers in patients with lym- phomas. leukemias and myeloma. J Natl Cancer Inst 38:741. I967. Berk Pl). Goldberg JD ct al: lncrcased incidence of acute leukemia in polycythcmia vcra associated with chlorambucil therapy. N Iingl J Med 304:441—447. I981, Bibbo M. Haenszel WM. Wied GL et al: A twenty-five- year follow-up study of women exposed to diethylslil- bestrol during pregnancy. N Engl J Med 298:763—767. I978. Biggar RJ. Henle W. Fleisher G: Primary Epstein-Barr virus infections in African infants. I. Decline of ma- ternal antibodies and time of infection. Int J Cancer 23:239—243. 1978. Bingham 5. William TR. Cole 'I‘J et al: Dietary fibre and regional large-bowel cancer mortality in Britain. BrJ Cancer 40:456—463. I979. Bithell JF. Stewart AM: Pre-natal irradiation and child- hood malignancy: A review of British data from the Oxford Survey. BrJ Cancer 31:27I—287. I975. Bjelke E: Dietary factors and the epidemiology of cancer of the stomach and large bowel. In Aktuelle Probleme der Klinischen Diatetik. Supplement Lu "Aktuelle Ernahrungsmcdicin,“ Stuttgart: George Thieme Verlag. [978. pp. [0—17. Blattner WA. et al: Geneology of cancer in a family. JAMA 241:259—261. I979. Blattner WA: Multiple Inycloina and Inacroglobuline- niia. In Cancer Epidemiology and Prevention (Schot- tcnfeld D and Fraumeni .ll" Jr. eds). Philadelphia: W.B. Saunders. I983. pp 795—Xl l. Blot WJ. Davies JE. Brown LM. Nordwall CW. Buiatti E. Ng A. Fraumeni JF Jr: Occupation and high risks of lung cancer in northeast Florida. Cancer 50:364—37I. I982. Blot WJ. Fraumeni JF Jr: Cancer among shipyard work- ers. Banbury Rpt 9:37—49, 19%|. Blot WJ. Fraumeni JF Jr: Geographic patterns of oral cancer in the United States: Etiologic implications, J Chron Dis 30:745—757. I977. Blot WJ. Fraumeni JF Jr. Stone BJ: Geographic pat- terns of breast cancer in the United States. J Natl Cancer Inst 59: I407—I4I I. [977. Blot WJ. Fraumeni JI’ Jr. Stone BJ. McKay FW: Geo- graphic patterns of large-bowel cancer in the United States. J Natl Cancer Inst 57: I225—I23I. I976. Blot WJ. Harrington JM. Toledo A et al: Lung cancer after employment in shipyards during World War II. N EnglJ Med 299:620—624. I978. Blot WJ. Morris LE. Strouhe R et al: Lung and phar- yngeal cancers in relation to shipyard employment in coastal Virginia. J Natl Cancer Inst 65:571—575. I980, Blot WJ. Winn DM. Fraumeni JF Jr: Oral cancer and mouthwash. J Natl Cancer Inst 70:251—253. 1983. Blumberg BS. London T: Hepatitis B virus: Patho- genesis and prevention of primary cancer of the liver. Cancer 50:2657—2665. 1982. Boice JD Jr. Greene MH. Killen JY et a1: Leukemia fol- lowing treatment with methy1~CCNU for gastroin» testinal cancers. N EnglJ Med 309:1079—1084. 1983. Boiee JD Jr. Land CE. Ionizing Radiation. In Cancer Epidemiology and Prevention (Schottent’eld I). Fraumeni JF Jr. eds). Philadelphia: W.B. Saunders 1982. pp 231—253, Boice JD Jr. Land CE. Shore RE et al: Risk of breast cancer following low—dose radiation exposure. Radiol- ogy 131:589—597. 1979. Boice JD Jr, Monson RR: Breast cancer in women after repeated fluoroscopic examinations of the chest. J Natl Cancer Inst 59:823—832. 1977. Brian DD. Tilley BC. Labarthc DR et al: Breast cancer in DES-exposed mothers. Absence of an association. Mayo Clin Proc 55:89—93. 1980. Brinton. LA. Blot WJ. Bcckcr JA ct al: A casc~contro| study of cancers of the nasal cavity and paranasal si- nuses. Am] lipidentiol ”91896—906. 1984. Brinton LA. Hoover RN et al: Menopausal estrogen use and risk of breast cancer. Cancer 47:2517—2521. 1981. Brinton LA. Hoover RN. S/klo M et al: ()ral contra- ceptives and breast cancer. Int J Epidemiol 11116322. 1982. Brinton LA. Stone BJ. Blot WJ. Fraumeni JF Jr: A death certificate analysis of nasal cancer among fur- niture workers in North Carolina. Cancer Res 373473-3474. 1977. Buell P: Relative impact of smoking and air pollution on lung cancer. Arch Environ Health 15:291—297. 1967. Cantor KP. Hoover R. Mason TJ. et al: Associations ot~ cancer mortality with halomethanes in drinking water. J Natl Cancer Inst 61 1979-985. 1978. Carroll KK. Khor HT: Dietary fat in relation to tu~ morigenesis. In Lipids and Tumors (Carroll KK. ed). Progr Biochem Pharmacol. Vol 10. Basel: Karger. 1975. pp 308—353. Case RAM. Hosker ME. McDonald DB et al: Tumors of the urinary bladder in workmen engaged in the manufacture and use of certain dyestulT intermediates in the British chemical industry. BrJ Ind Med 11:75—104. 1954. Ccdcrlof R. Fribcrg L. Hrubec Z et al: The relationship of smoking and some social covariablcs to mortality and cancer mortality. A ten-year follow-up on a prob- ability sample of 55 .000 Swedish subjects age 18 to 69. Stockholm: Dept of Environmental Hygiene. The Karolinska Institute. 1975. Centers for Disease Control: Long-term oral contra- ceptive use and the risk of breast cancer. JAMA 249:1591—1595. 1983. Centers for Disease Control: Oral contraceptive use and the risk of ovarian cancer. JAMA 249: 1596—1599. 1983. Centers for Disease Control: Oral contraceptive use and the risk of endometrial cancer. JAMA 249:1600— 1604. 1983. Choi NW, Sehuman I.M.Gu11en WH: Epidemiology of primary central nervous system neoplasms: 11. Case control study. Am J Epidemiol 91:467—485. 1970. Clemmesen J: On the etiology of some human cancers. J Natl Cancer Inst 12:1—21. 1951. Coldman AJ. Elwood JM. Gallagher RP: Sports activi- ties and risk of testicular cancer. BrJ of Cancer 46:749—756. 1982. Cole P. Hoover R. Friedell GH: Occupation and cancer ofthe lower urinary tract. Cancer 29:1250—1260. 1972. Correa P: Precursors of gastric and esophageal cancer. Cancer 50:2554—2565. 1982. Correa P. O'Conor GT: Epidemiologie patterns of Hodgkin‘s disease. Int J Cancer 8: 192—201. 1971. Correa P. Pickle LW. Fontham ETH. Johnson WD: Pre- liminary report on a case«control study of cancers of the lung. stomach. and pancreas in southern Loui- siana. In Progress on Joint Environmental and Oc- cupational Cancer Studies. September 9—11. 1981. Proceedings of the Second NCI/EPA/NIOSH Collab- orative Workshop. Rockville. Maryland. printed April. 1982. pp 67—88. Court Brown WM. Doll R: Mortality from cancer and other causes after radiotherapy for ankylosing spon- dylitis. Br MedJ 221327—1332. 1965. Cramer DW. Welch WR. Scully RE et a1: Ovarian can- cer and tale. a case-control study. Cancer 50:372—376. 1982, Crump KS. Guess HA: Drinking water and cancer: Re~ view of recent epidemiological findings and assess- ment of risks. Ann Rev Public Health 3:339—357. 1982. Cuello C. Correa P. Hacnszel W. et al: Gastric cancer in Colombia. J Natl Cancer Inst 57: 1015—1020. 1976. Cutler SJ. Young JL Jr (eds): Third National Cancer Survey: Incidence Data. Natl Cancer Inst Monogr 41. 1975. da Silva Horta J. da Motta LC. Tavares MH: Thorium dioxide effects in man. Environ Res 8:131—159. 1974. Day NE. Munoz N: Esophagus. In Cancer Epidemiolo- gy and Prevention (Schottenfeld D. Fraumeni JF Jr. edst. Philadelphia: W.B. Saunders. 1982. pp. 596—623. Dean G. MacLennan R. McLoushlin H. Shelley E: Causes of death of blue-collar workers at a Dublin brewery. 1954—73. BrJ Cancer 40(4): 581-589. 1979. 129 Decoufle P: Cancer risks associated with employment in the leather and leather products industry. Arch of Envir Health 34:33—37. 1979. De Jong UW. Breslow N. Goh Ewe HongJ et al: Aetiological factors in oesophageal cancer in Sin- gapore Chinese. Int J Cancer 13:291—303. 1974. DES Task Force Summary Report. NIH Pub] No. 81-1688. Bethesda. Md.: U.S. Dept of Health & Human Serv. September 21. 1978. Dchsa SS. Silverman DT: Cancer incidence and mor- tality trends in the United States: 1935—74. J Natl Cancer Inst 60:545—571. 1978. de Waard F: Breast cancer incidence and nutritional status with particular reference to body weight and height. Cancer Res 353351—3356. 1975. de Waard F: Uterine corpus. In Cancer Epidemiology and Prevention (Schottenfeld D and Fraumeni JF Jr. eds). Philadelphia: W.B. Saunders. 1982. pp 901—908. Doll R. l’eto R: Mortality in relation to smoking: 20 years' observations on male British doctors. Br Med J 2:1525—1536. I976. Doll R. I’eto R: The causes of cancer: quantitative esti- mates of avoidable risks of cancer in the United States today. J Natl Cancer Inst 66:1193—1308. 1981. Dunn JE Jr. Levin EA. Linden G et al: Skin cancer as a cause of death. Calif Med [02:361—363. 1965. Economic Research Service. U.S. Department of Agri- culture. Tobacco Outlook and Situation. TS-183. March 1983. USDA. Washington DC. Ehrich M. Aswell JE. Wilkins TI): Alteration ofthc mu- tagenicity of human fecal extracts by hepatic micro- somal en/ymes. J ’l'oxicol Environ Health 7: 107—1 15. 1981. Elwood JM. Cole P. Rothman KJ. Kaplan SD: Epidemi» ology ofendometrial cancer. J Natl Cancer Inst 59(4):]055—1060. 1977. Enstrom JE: Cancer and total mortality among active Mormons. Cancer 42: 1943—1971. 1978. Ernster VL. Selvin S et al: Prostate cancer: mortality and incidence rates by race and social class. Am J Epidemiol 1072311—320. 1978. Falk H: Liver. In Cancer Epidemiology and Prevention (Schottenfeld D and Fraumeni JF Jr. eds). Phila» delphia: W.B. Saunders. 1982. pp 668—682. Fears TR. Scotto J: Changes in skin cancer morbidity between 1971—72 and 1977—78. J Natl Cancer Inst 69:365—370. 1982. Fears TR. Scotto J: Estimating increases in skin cancer morbidity due to increases in ultraviolet radiation ex- posure. Cancer Investigation Vol I. No. 2. 1983. Findlay GM: Ultraviolet light and skin cancer. Lancet 2:1070—1073. 1928. Fraumeni JF Jr: Persons at High Risk of Cancer. New York: Academic Press. 1975. 130 Fraumeni JF Jr. Blot WJ: Lung and pleura. In Cancer Epidemiology and Prevention (Schottenfeld D and Fraumeni JF Jr. eds). Philadelphia: WB Saunders. 1982. Fraumeni JF Jr. Hoover R: lmmunosurveillance and cancer: Epidemiologic observations. Natl Cancer Inst Monogr 47:121—126. 1977. Fraumeni JF Jr. Kantor AF: Biliary Tract. In Cancer Epidemiology and Prevention (Schottenfeld D and Fraumeni JF Jr. eds). Philadelphia: W.B. Saunders. 1982. pp 683—691. Gallo RC: Editorial: Gallo on T-cell leukemia-lympho- ma virus. The Lancet 2: 1083. November 13. 1982. Gatti RA. Good RA: Occurrence of malignancy in im- munodeficiency diseases. Cancer 28:89—98. 1971. Geleprin A. Moses V. Fox G: Nitrate in water supplies and cancer. Illinois MedJ 149:251-253. 1976. Gilbert JB: Tumors of the testis following mumps or— chitis: Case report and review of 24 cases. J Urol 51:296—300. 1944. Glass AG. Fraumeni JF Jr: Epidemiology of bone can» cer in children. J Natl Cancer Inst 44: 187—199. 1970. Gobar AH: L‘abus des drogucs en Afghanistan. Bull Stupefiants 2821—12. 1976. Gold E. Gordis L. Tonascia J. Szklo M: Risk factors for brain tumors in children. Am J Epidemiol 1091309— 319. 1979. Graham S. Dayal H. Rohrer T et al: Dentition. diet. to- bacco. and alcohol in the epidemiology of oral cancer. J Natl Cancer Inst 59:1611—1618. 1977. Graham S. Dayal H. Swanson M et al: Diet in the epi- demiology ofcancer of the colon and rectum. J Natl Cancer Inst 61:709—714. 1978. Graham S. Gibson R, West 1) et al: Epidemiology of cancer of the testis in Upstate New York. J Natl Can- cerlnst5811255—1261. 1977. Graham S. Haughey B. Marshall J et al: Diet in the epi- demiology ofcarcinoma of the prostate gland. J Natl Cancer Inst 70:687—692. 1983. Greene MH: Non-Hodgkin's Lymphoma and Mycosis Fungoides. In Cancer Epidemiology and Prevention ISchottenfeld I) and Fraumeni JF Jr. eds). Phila- delphia: WB. Saunders. 1982. pp 754—778. Greene MH. Boice JD Jr et al: Acute non-lymphocytic leukemia after therapy with alkylating agents for ovarian cancer: A study of five randomized trials. N Engl J Med 307:1416—1421. 1982. Greene MH. Clark WH Jr. Tucker MA et al: Precursor naevi in cutaneous malignant melanoma: A propOsed nomenclature. Lancet. November 8. 1980. pp. 1024. Greene MH. Goldin LR. Clark WH et al: Familial cuta- neous malignant melanoma—an autosomal dominant trait possibly linked to the Rh locus. Proc Natl Acad Sci 80:6071—6075. 1983. Greene MH. Hoover RN. Eek RL eta]: Cancer mor» tality among printing plant workers. Environ Res 20:66—73. 1979. Greene MH. Young RC et al: Evidence ofa treatment dose-response in acute nonlymphocytic leukemias which occur after therapy of non-Hodgkin‘s lympho- ma. Cancer Res 431891—1898. I983. Greenwald P: Prostate. In Cancer Epidemiology and Prevention (Schottenfeld D and Fraumeni JF Jr. eds). Philadelphia: W.B. Saunders. I982. pp 938-946. Greenwald P. Korns. RF. Nasca PI. ct al: Cancer in United States Jews. Cancer Res 35:3507—3512. I975. Grufferman S: Hodgkin‘s disease. In Cancer Epidemiol- ogy and Prevention (Schottenfeld D and Fraumeni JF Jr. eds). Philadelphia: W.B. Saunders. I982. pp 564—582. Grufferman S. Duong T. Cole P: Occupation and Hodg- kin's disease. J Natl Cancer Inst 57: I 193—] 195. I976. Haenszel W. Kurihara M: Studies of Japanese migrants. I. Mortality from cancer and other diseases among Japanese in the United States. J Natl Cancer Inst 40:43—68. 1968. Haenszel W. Kurihara M. Segi M el al: Stomach cancer among Japanese in Hawaii. J Natl Cancer Inst 49:969—988. I972. Hammond EC: Smoking habits and air pollution in rela- tion to lung cancer. In Environmental Factors in Res- piratory Disease (Lee DHK. ed). New York: Aca- demic Press. 1972. pp 177—198. Hammond EC: Smoking in relation to the death rates of I million men and women. In Epidemiological Ap- proaches to the Study of Cancer and Chronic Dis- eases (Haenszel W, ed). Natl Cancer Inst Monogr I9. I966. pp 127—207. Hammond EC. Selikoff IJ. Lawther PL et al: Inhalation of Benzpyrene and cancer in man. Ann NY Acad Sci 27I1116—124.I97o. Hammond EC. Selikoff IJ. Scidman H: Asbestos ex— posure, cigarette smoking and death rates. Ann NY Acad Sci 330:473—490. 1979. Harris RH. Page 'I‘. Reiches NA: Carcinogenic hazards of organic chemicals in drinking water. In Incidence of Cancer in Humans (Hiatt HH. Watson JD. Winsten JA. eds). Cold Spring Harbor. New York: CSH Labo- ratory. I977. pp. 309—330. Harris RWC. Brinton LA. Cowdell RH et al: Character- istics of women with dysplasia or carcinoma in sir/1 of the cervix uteri. Br J Cancer 42:359—369. I980. Hartge P. Hoover R. West DW et al: Coffee drinking and risk of bladder cancer. J Natl Cancer Inst 7()(6):IOZI—I026. I983. Hartge P. Leshcr L. McGowan L. Hoover R: Talc and ovarian cancer (letter). JAMA 250( I4):l844. I983. Hecht SS. Chen CB. ()hmori T. Hoffman D: Com- parative carcinogenicity in F344 rats ofthe tobacco- specific nitrosamines. N.nitrosonornicotine and 4-(N- methyl-N-nitrosamino)»l-(3-pyridyl)—I-butanone. Cancer Res 40:298-302. I980. Hempelmann LH. Hall WJ. Phillips M. Ames WR: Neoplasms in persons treated with X-rays in infancy: Fourth survey in 20 years. J Natl Cancer Inst 55:5I9—530. I975. Henderson BE. Benton B. Jing J, Yu MC. Pike MC: Risk factors for cancer ofthe testis in young men. Int J Cancer 23:598—602. I979. Henderson WJ. Joslin CAF. Turnvull AC et al: Talc and carcinoma of the ovary and cervix. J ()bstet Gynecol Br Comm 78:266—272. I971. Herhst AL. Ulfelder H. Poskanzer DC: Adenocarcino— ma of the vagina: Association of maternal stilbestrol therapy with tumor appearance in young women. N Engl] Med 284:878—881. 1971. Hildick—Smith GY: The biology of talc. BrJ Industr Med 33:2I7—229. I976. Holly EA. Weiss NS. LiITJM: Cutaneous melanoma in relation to exogenous hormones and reproductive factors. J Natl Cancer Inst 70:827—831. I983. Hoover R: Effects of drugs—immunosuppression. In Origins of Human Cancer (Hiatl HH. Watson JD. Winsten JA. cdst. Cold Spring Harbor. New York: Proc of CSH Conferences on Cell Proliferation. I977. pp 369—379. Hoover R. Fraumeni JF Jr: Drugs in clinical use which cause cancer. J Clin Pharmacol 15:16—23. 1975. Hoover R. Fraumeni JF Jr: Drug-induced cancer. Can- cer 47:1071—1080. I98]. Hoover R. Fraumeni JF Jr: Risk ol‘cancer in renal transplant recipients. Lancet 2:55. I973. Hoover R. Gray LA. Cole P et al: Menopausal es- trogens and breast cancer. N Engl J Med 295:401—405. 1976. Hoover RN. Strasser I’H ct al: Artificial sweeteners and human bladder cancer. Lancet. April 19. I980. pp. 837—840. Horm JW. Asire AJ: Changes in lung cancer incidence and mortality rates among Americans: 1969—78. J Natl Cancer Inst 69:833—837. I982. Hounam RF. Williams J: Levels of airborne dust in fur- niture-making factories in the High Wycombe area. BrJ Ind Med 3|:I—9. I974. Hulka BS. Chamblcss LE et al: Breast cancer and es- trogen replacement therapy. Am J ()hstet Gynecol I43zo38—644. I982. Hulka BS. Chambless. LE. Kaufman [)0 ct al: Protec— tion against endometrial carcinoma by combination— producl oral contraceptives. JAMA 247(4):47S—477. I982. Hyde JN: ()n the influence of light in the production of cancer of the skin. Am J Med Sci 13111-22. 1906. International Agency for Research on Cancer (IARCI monographs on the evaluation of the carcinogenic risk of chemicals to humans. Lyons. France: Some anti-thyroid and related substances. nitrofurans and industrial chemicals. Vol 7. 1974. Asbestos. Vol [4. 1977. Some metals and metallic compounds. Vol 23. 1980. Wood. leather. and some associated industries. Vol 25. 1981. Chemicals. industrial processes and industries associ- ated with cancer in humans. Vols. 1—20. Supple- ment 4. 1982. Ichimaru M. Ishimaru T. Mikami M et al: Multiple my« eloma among atomic bomb survivors in Hiroshima and Nagasaki. 1950—1976: Relationship to radiation dose absorbed by marrow. J Natl Cancer Inst 69:323. 1982. Jain M. Cook GM. Davis FG. Grace MG. Howe GR. Miller AB: A case-control study of diet and colo-rec- tal cancer. Int J Cancer 26:757—768. 1980. Jayant K. Balakrishnan V. Sanghvi LD. Jussawalla DJ: Quantification of the role of smoking and chewing to— bacco in oral. pharyngeal. and esophageal cancers. BrJ Cancer 35:232—235. 1977. Jensen ()M: Cancer morbidity and causes of death among Danish brewery workers. Int J Cancer 23:454—463. 1979. Jick H. Walker AM et al: Replacement estrogens and breast cancer. AmJ Epidemiol 112:586—594. 1980. Jick H. Watkins RN. Hunter JR et al: Replacement es- trogens and cndometrial cancer. New Engl J Med 300(512218—222. 1979. Johnson FL. Feagler JR. Lerner KG et al: Association of androgenic anabolic steroid therapy with develop- ment of hepatocellular carcinoma. Lancet 2:1273— 1276. I972. Jussawalla DJ: Epidemiological assessment of aetiology of oesophageal cancer in greater Bombay. Interna- tional Seminar on Epidemiology of Oesophageal Can- cer. Bangalore. India. Monogr 1. November 1971. pp 20—30. Kato H. Tillotson J et al: Epidemiologic studies of coro» nary heart disease and stroke in Japanese men living in Japan. Hawaii and California: Serum lipids and diet. AmJ Epidemiol 97:373-385. 1973. Kelsey JL: A review ofthe epidemiology of human breast cancer. Epidem Reviews 1:74—109. I979. Kelsey JL. Fischer DB. Holford TR et al: Exogenous estrogens and other factors in the epidemiology of breast cancer. J Natl Cancer Inst 67:327—333. 1981. 132 Kew MC.(Jcar AJ. Baumgarten I et a1: Histocom- patibility antigens in patients with hepatocellular car- cinoma and their relationship to chronic hepatitis B virus infection in these patients. Gastroenterology 77:537—539. 1979. Kinlen LJ. Shiel AGR. Peto J ct al: A collaborative study of cancer in patients who have received immu- nosuppressive therapy. Br Med J 2: 1461—1466. 1979. Kirsner JB: Introduction: Inflammatory bowel dis- ease—consideration of etiology and pathogenesis. In Colorectal Cancer: Prevention. Epidemiology and Screening (Winawer SJ. Schottcnfeld D. Sherlock P. eds.) New York: Raven Press. 1980. pp 319—323. Knox EG: Foods and disease. BrJ Prev Soc Med 31:71—80. 1977. Kolonel LN. Hankin JH. Lee J et al: Nutrient intakes in relation to cancer incidence in Hawaii. BrJ Cancer 44:332—339. 1981. Kolonel L. Winkelstein W Jr: Cadmium and prostate carcinoma. Lancet 2:566—567. 1977. Knudson AG Jr. Meadows AT: Developmental genetics of neuroblastoma. J Natl Cancer Inst 57:675—682. 1976. Kraybill HF et al: Evaluation of public health aspects of carcinogenic/mutagenic biorel‘ractories in drinking water. Preventive Med 9:212—218. 1980. Lew EA. Garfinkel L: Variations in mortality by weight among 750.000 men and women. J Chronic Dis 32:563—576. 1979. Li FP: Cancers in children. In Cancer Epidemiology and Prevention (Schottenfeld D and Fraumeni JF Jr. eds). Philadelphia:W.B. Saunders. 1982. pp 1012—1024. Li FP. Connelly RR. Myers M: Improved survival rates among testicular cancer patients in the US. JAMA 247(6):825—826. February 1982. Li FP. Fraumeni JF Jr: Prospective study of a family cancer syndrome. JAMA 247:19. 1982. Li FP. Fraumeni JF Jr: Rhabdomyosarcoma in children: Epidemiologic study and identification of a familial cancer syndrome. J Natl Cancer Inst 43: 1365—1373. 1969. Li FP. Fraumeni JF Jr: Soft-tissues sarcoma. breast cancer. and other neoplasms: A familial syndrome. Ann Int Med 71:747—752. 1969. Li FP. Fraumeni JF Jr. Mantel N et al: Cancer mortality among chemists. J Natl Cancer Inst 43:1159—1164. 1969. Lindqvist C: Risk factors in lip cancer: A questionnaire survey. AmJ Epidemiol 1092521—530. 1979. Lloyd JW. Decoufle P. Salvin LF: Unusual mortality experience of printing pressman. J Occup Med 19(812543—550. I977. Lundin FE Jr.. Wagoner JK. Archer VE: Radon Daughter Exposure and Respiratory Cancer: Quan- titative and Temporal Aspects. NI()SH and NIEHS Joint Monogr No. 1. Springfield. Virginia: NTIS. 1971. Lyon JL. Gardner JW. West DW: Cancer risk and life- style: Cancer among Mormons from 1967 to 1975. In Banbury Report 4: Cancer Incidence in Defined Pop- ulations (Cairns J. Lyon JL. Skolnick M. eds). Cold Spring Harbor. New York: CSH Laboratory. I981. pp 3—30. Lyon JL. Sorenson AW: Colon cancer in a low-risk population. Am J Clin Nutr 31:527—230. 1978. Mack TM. Paganini-Hill A: Epidemiology of pancreas cancer in Los Angeles. Cancer 47: 1474—1481. 1981. MacDonald PC. Siiteri PK: The relationship between the extraglandular production of estrone and the oc- currence of endomctrial neoplasia. Gynecol ()ncol 2:259—263. 1974. MacLennan R. Jensen ()M. Mosbech J. Vuori H: Diet. transit time, stool weight. and colon cancer in two Scandinavian populations. Am J Clin Nutr 3118239— 242. 1978. MacMahon B. Cole P. Lin TM et al: Age at first birth and cancer ofthe breast. A summary of an interna- tional study. Bull WHO 43:209—221. 1970. MacMahon B. Yen S. Trichopoulos D el al: Coffee and cancer ofthe pancreas. New Engl J Med 304:630— 633. 1981. Mahboubi E. Sayed GM: ()ral cavity and pharynx. In Cancer Epidemiology and Prevention (Schottenfeld D and Fraumeni JF Jr. eds). Philadelphia: W.B. Saun- ders. 1982. pp 583—595. Malik MOA. Hidaytalla A. Daoud EH et al: Superficial cancer in the Sudan—a study of 1225 primary malig- nant superficial tumours. Br J Cancer 30:355—364. 1974. Malik MOA. Zaki EL. Din ZA. El Masri SH: Cancer of the alimentary tract in the Sudan—a study of 546 cases. Cancer 37:2533—2542. 1976. Maneuso TF. Brennan MJ: Epidemiological considera< tions of cancer of the gallbladder. bile ducts and sali- vary glands in the rubber industry. J ()ccup Med 12:333—341. 1970. Mason TJ: Cancer mortality in U.S. counties with plas- tics and related industries. Environ Health Perspect 11:79—84. 1975. Mason TJ. McKay FW. Hoover R. Blot WJ, Fraumeni JF Jr: Atlas of Cancer Mortality in U.S. Counties 1950—1969. DHEW Publ No. (NIH) 75-780. Washing- ton. D.C.: U.S. Govt Print Off. 1975. Matanoski GM. Sartwell 1’. Elliott EA: Hodgkin's dis- ease and mortality among physicians. Lancet ii:926—927. 1975. Matanoski GM. Seltser R, Sartwell PE et al: The cur- rent mortality rates of radiologists and other physi» cian specialists: Specific causes of death. Am J Epi- demiol lOll3):l99—210. I975. McKay FW. Hanson MR, Miller RW: Cancer Mortality in the United States: 1950—1977. DHHS Publ No. (NIH) 82-2435. Natl Cancer Inst Monogr 59. Wash- ington. D.C.: U.S. Govt Print Off. 1982. McKusick VA: Mendelian inheritance in man. Catalogs of autosomal dominant. autosomal recessive. and X- Iinked phenotypes. Baltimore: The Johns Hopkins University Press. 1978. PP. 605—606. McLaughlin JK. Blot WJ. Mandel JS et al: Etiology of cancer of the renal pelvis. J Natl Cancer Inst 71:287— 291. I983. McLaughlin JK, Mandel JS. Blot WJ et al: A popula- tion-based case—control study of renal cell carcinoma. J. Natl Cancer Inst 72(2):275—284. 1984. McMichael AJ. Hartshorne JM: Mortality risks in Aus- tralian men in occupational groups. 1968—78. Med J of Australia 1:253—256. 1982. Medical Research Council (U.K.): The Hazards to Man of Nuclear and Allied Radiations. London: H.M. Sta- tionery Office. 1956. Mettlin C. Graham S. Priore R et al: Diet and cancer of the esophagus. Nutrition and Cancer 22143—147. I981. Mettlin C. Graham S. Swanson M: Vitamin A and lung cancer. J Natl Cancer Inst 62:1435—1438. 1979. Milham S: Occupational mortality in Washington State. 1950—1971.DHEW Publ No. (NIOSH) 76-175—C. Washington. DC: U.S. Govt Print Off. 1976. Miller AB. Howe GR et al: Food items and food groups as risk factors in a case—control study ofdiet and co- lorectal cancer. Int J Cancer 32:155—161. I983. Miller RW: Cancer epidemics in the People‘s Republic of China. J Natl Cancer Inst 60: I 195. 1978. Miller RW. Epidemiology of leukemia. In Modern Trends in Human Leukemia III. New York: Springer— Verlag. 1979. pp 37-40. Miller RW: Ethnic ditTerences in cancer occurrence: Genetic and environmental influences with particular reference to neuroblastoma. In Genetics of Human Cancer (Mulvihill J]. Miller RW. Fraumeni JF Jr. eds). New York: Raven Press. 1977. pp 1—14: Discussion pp 39—41. Miller RW: Genetic and familial factors. In Medical ()n- cology. New York: MacMillan. in press. Miller RW. Beebe GW: Infectious mononucleosis and the empirical risk of cancer. J Natl Cancer Inst 50:315—321. 1973. Miller RW. Fraumeni JF Jr. Hill JA et al: Neuroblas— toma: Epidemiologic approach to its origin. AmJ Dis Child 115:253—261. 1968. 133 Moertel CG. Hagedorn AB: Leukemia or lymphoma and coexistent primary malignant lesions: A review of the literature and study of 120 cases. Blood |2:788. 1957. Morrison AS: Cryptorchidism. hernia. and cancer of the testis. J Natl Cancer Inst 56:731—733. 1976. Morrison AS. Buring JE: Artificial sweeteners and can- cer of the lower urinary tract. N Engl J Med 302:537— 541. 1980. Morrison AS. Buring JE. Verhock WU et al: Coffee drinking and cancer of the lower urinary tract. J Natl Cancer Inst 68:91—94. I982. Morrison AS. Cole P: Epidemiology of bladder cancer. Urol Clin North Am 3: 13—29. 1976. Morrison AS. Cole P: Urinary tract. In Cancer Epide- miology and Prevention (Schottenfeld I) and Frau- meni .IF .lr. eds). Philadelphia: W.B. Saunders. 1982. pp 925—937. Myers MH. Hankey Bl“: Cancer Patient Survival Expe- rience: Trends in Survival 1960—63 to 1970—73. DHHS Publ No. (NIH) 80-2148. Bethesda Md.: Natl Inst of Health. June 1980. NAS/NRC: The Biological Effects ofAtomic Radiation: Summary Reports. Washington. DC: NAS/NRC. I956. National Toxicology Program. DH HS. USI’HS. Rc~ search Triangle Park. NC. Tech. Report No. 289. I984 (draft). Newhouse ML. Berry (J. Wagner JC et al: A study of the mortality of female asbestos workers. Br J Industr Med 29:134-141. 1972. Nomura A: Stomach. In Cancer Epidemiology and Pre- vention (Schottenfeld D and Fraumeni JF Jr. eds). Philadelphia: W.B. Saunders. I982. pp 624—637. ()lin GR. Ahlbom A; The cancer mortality among Swedish chemists graduated during three decades. A comparison with the general population and with a cohort of architects. Environ Res 22:154—161. I980. Palmer S. Bakshi K: Diet. nutrition. and cancer. Inter- im dietary guidelines. J Natl Cancer Inst 70: l 151— 1170. 1983. Peto R. Doll R. Buckley JD. Sporn MB: Can dietary beta-carotene materially reduce human cancer rates? Nature 290:201—208. 1981. Phillips RL et al: Mortality among California Seventh- Day Adventists for selected cancer sites. J Natl Can» cer Inst 65:1097—1107. 1980. Pickle LW. Greene MH, Ziegler RU et al: Colorectal cancer in rural Nebraska. Cancer Res 44:363—369. 1984. Pike MC. Krailo MD. Henderson BE et al: Breast can- cer in young women and use of oral contraceptives: Possible modifying effect of formulation and age at use. Lancet 8359:926—929. I983. Poiesz BJ. Ruseetti FW. Gazdar AF et al: Detection and isolationoftype C retrovirus particles from fresh and cultured lymphocytes of a patient with cutaneous T- cell lymphoma. Proc Natl Acad Sci 77:7415—7419. 1980. Pottern LM. Ziegler RG et al: Esophageal cancer among black men in Washington. DC: 1. Alcohol. tobacco. and other risk factors: 11. Role of nutrition. J Natl Cancer Inst 67:777—783: 1199—1206. 1981. Prates MD. Torres F0: A cancer survey in Lourenco Marques. Portuguese East Africa. J Natl Cancer Inst 35:729—757. 1965. Purtilo DT: Opportunistic non-Hodgkin's lymphoma in X-linked reessive immunodeficiency and lympho- proliferative syndromes. Semin Oncol 41:335—343. 1977. Radl .l. Sepers JM. Skvaril F ct al: Immunoglohulin pat- terns in humans over 95 years of age. Clin Exp Im- munol 22:84—90. 1975. Rehn L: Blasengeschulste bei Fuchsin-Arbeitern. Arch Klin Chir 50:588—600. 1895. Ries LG. Pollack ES. Young JL Jr: Cancer patient sur- vival: Surveillance. Epidemiology, and End Results Program. 1973—79. J Natl Cancer Inst 70:693—707. 1983. Ritchie JK. Allan RN. Macartney .I et al: Biliary tract carcinoma associated with ulcerative colitis. Quart J Med 43:263—279. 1974. Rosdahl N. Larsen S. Clemmesen J: Hodgkin's disease in patients with previous infectious mononucleosis: 30 years' experience. Br Med J 2(913):253—256. May 4. 1974. Ross RK. Paganini-Hill A ct al: A case-control study of menopausal estrogen therapy and breast cancer. JAMA 243:1635-1639. 1980. Rothman K. Keller A: The effect ofjoint exposure to alcohol and tobacco on the risk of cancer of the mouth and pharynx. J Chron Dis 25:711—716. I972. Rotkin ID: Epidemiology of cancer of the cervix. Ill. Sexual characteristics of a cervical cancer popula- tion. Am J Public Health 57:815—829. 1967. Schindler AE. Ebert A. Friedrick E: Conversion of an- drostenedione to estrone by human fat tissue. J Clin Endocrinol Metab 35:627—630. I972. Schoenberg BS: Nervous System. In Cancer Epidemi- ology and Prevention (Sehottenfeld D and Fraumeni JFJr. eds). Philadelphia: W.B. Saunders. 1982. pp 968—983. Schottenfeld D. Warshauer ME: Testis. In Cancer Epi— demiology and Prevention (Schottenfeld D and Fraumeni JF Jr. eds). Philadelphia: W.B. Saunders. I982. pp 947—957. Schottenfeld D. Winawer SJ: Large Intestine. In Cancer Epidemiology and Prevention (Schottenfeld D and Fraumeni JF Jr, eds). Philadelphia: W.B. Saunders. 1982. pp. 703—727. Schreier HA. Sherry N. Shaughnessy E: Lead poison- ing & brain tumors in children: A report of 2 cases. Ann Neurol l2599—600. 1977. Scotto J. Fears TR, Fraumeni JF Jr: Incidence of Non- melanoma Skin Cancer in the United States. DHHS Publ No. (NIH) 82—2433. Washington. DC: U.S. Govt Print Off. I98I. Scotto J. Fears TR. Fraumeni JF Jr: Solar radiation. In Cancer Epidemiology and Prevention (Schottenfeld D. Fraumeni JF .lr. eds). Philadelphia: W.B. Saun- ders. 1982. pp 254-274. Scotto J. Fears TR. Gori GB: Measurements of Ultra- violet Radiation in the United States and Com- parisons with Skin Cancer Data. DH EW Publ No. (NIH) 76-1029. Washington. DC: U.S. Govt Print 011‘. I976. pp 3.l.—3.IO. Scotto J. Fraumeni JF Jr: Skin (other than melanoma). In Cancer Epidemiology and Prevention (Scottenfeld D. Fraumeni JF .Ir. eds). Philadelphia: W.B. Saun- ders. I982. pp 996—110]. Selikoff IJ. Hammond EC. Seidman H: Mortality expe- rience of insulation workers in the United States and Canada. I943—l976. Ann NY Acad Sci 330:91—116. I979. Shekelle RB. Liu S. Raynor WJ et al: Dietary vitamin A and risk of cancer in the Western Electric study. Lancet. November I8. I98]. Shimkin MB: Contrary to Nature. DHEW Publ No. (NIH) 76-720. Washington. D.C.: U.S. Govt Print ()ff. I977. Shy CM. Struba RJ: Air and water pollution. In Cancer Epidemiology and Prevention tSchottcnfeld D and Fraumeni JF Jr. eds). Philadelphia: W. B. Saunders. I982. pp 336—363. Silverberg E, Lubera JA: Cancer Statistics I983. Ameri- can Cancer Society. CA 33 (adapted from SEER data). Silverman DT, Hoover RN. Albert S et al: Occupation and cancer of the lower urinary tract in Detroit. J Natl Cancer Inst 70:237—245. I983. Sinnhuber R0. Wales JH. Ayres JL et al: Dietary fac- tors and hepatoma in rainbow trout (Salmo Gardnci- ii). I. Aflatoxins in vegetable protein feedstuffs. J Natl Cancer Inst 4I27l I—7I8. I968. Smith EM: Epidemiology of oral and pharyngeal can- cers in the United States: Review of recent literature. J Natl Cancer Inst 63:| I89—I I98. I979. Smith PG. Doll R: Mortality among patients with an- kylosing spondylitis after a single treatment course with X-rays. Br Med J (Clin Res) 284163l4): 449—460. Feb I3. I982. Smith PG. Kinlen LJ. Doll R: Hodgkin's disease mor« tality among physicians. Lancet ii1525. I974. Sporn MB: Prevention of epithelial cancer by vitamin A and its synthetic analogs (retinoids). In Origins of Human Cancer (Hiatt HH. Watson JD, Winsten JA. eds). Cold Spring Harbor. New York: CSH Laborato- ry. 1977. pp SDI—807. Staszewski J: Migrant studies in alimentary tract can- cer. In Current Problems in the Epidemiology of Can» cer and Lymphomas (Grandmann E. Tulinius H eds). New York: Springer-Verlag. I972. Stern RS. Thihodeau LA. Kleinerman RA et al: Risk of cutaneous carcinoma in patients treated with oral methoxsalen photochemotherapy for psoriasis. N Engl J Med 300:809—813. I979. Stolley PD. Hibberd PL: Drugs. In Cancer Epidemiolo- gy and Prevention (Schottenfeld D and Fraumeni JF Jr. eds). Philadelphia: W.B. Saunders. 1982. pp 304-317. Stott H. Fox W et al: Acute leukaemia after busulphan. Br Med J 221513—1517. I977. Swerdlow AJ. Stiller CA. Kinnier—Wilson LM: Prenatal factors in the aetiology of testicular cancer: An epi» demiological study ofchildhood testicular cancer deaths in Great Britain. 1953—73. J Epid & Comm Health 36:96—l0l. I982. Szmuness W. Stevens CE. lkram H ct al: Prevalence of hepatitis B virus infection and hepatocellular car- cinoma in Chinese-Americans. J lnfec Dis I37:822— 829. I978. Thomas DB: Non-contraceptive exogenous estrogens and risk of breast cancer: A review. In Breast Cancer Research and Treatment. The Hague: Martinus Nij- hoff Pub. I982. Vol 2. pp 203—2I I. Thomas TL. Decoufle P: Mortality among workers em- ployed in the pharmaceutical industry: A preliminary investigation. J()M 2|:619—623. I979. Thomas TL. Waxweiler RJ. Moure-Eraso R et al: Mor- tality patterns among workers in three Texas oil refin- eries. J()M 24135—141. I982. Thomas TL. White DW. Moure‘Eraso R et al: Brain cancer among ()CAW members in three Texas oil re- fineries. Ann NY Acad Sci 38I:I2()—|29. I982. Tseng WP: Effects and dose-response relationships of skin cancer and hlack~foot disease with arsenic. In- ternational Conference on Arsenic. Fort Lauderdale. Florida. I976. Tseng WP. Chu HM. How SW. ct al: Prevalence of skin cancer in an endemic area of chronic arsenicism in Taiwan. J Natl Cancer Inst 40:453—463. I968. Tuyns AJ: Alcohol. In Cancer Epidemiology and Pre- vention (Schottenfeld D and Fraumeni JF Jr. eds). Philadelphia: W.B. Saunders. I982. pp 293—303. Tuyns AJ: Epidemiology of alcohol and cancer. Cancer Res 39:2840—2843. 1979. Tuyns AJ. chuignot G. Abhatucci JS: Oesophageal cancer and alcohol consumption. Importance of type of beverage. Int J Cancer 23:443—447. 1979. I35 'l‘uyns AJ. Pequignot G. Jensen OM: Le cancer de I‘ocsophage en IIIe-et-Vilaine en fonction des niveaux de consommation d‘alcool et de tabac. Des risques qui se multiplient. Bull Cancer 64:45—60. I977. 'I‘uyns AJ. Pequignot G. Jensen OM: Role of diet. alco» hol and tobacco in oesophageal cancer as illustrated by two contrasting high incidence areas in North Iran and west of Africa. In Frontiers of Gastrointestinal Research. Vol 4. Gastrointestinal Cancer: Advances in Basic Research (Rozen P. Eidelman S. Gilat T. eds). Basel: Karger. I979. pp 101—] I(). ’I‘yroler HA. Andjelkovic D. Harris R et al: Chronic dis» ease in the rubber industry. Environ Health Persp I7zl3—20. 1976. Urbach F: Geographic distribution of skin cancer. J Surg Oncol 31219—234. I97I. U.S. Department of Health and Human Services: The Health Consequences of Smoking: A Report ot‘the Surgeon General. DHHS Pub] No. tPHSI 82-50I79. Washington, D.C.: U.S. Gov Print Off. I982. Vianna NJ. Greenwald P. Davies JNP: Extended epi- demic of Hodgkin‘s disease in high school students. Lancet i:lZ()9—IZII. I97]. Vianna NJ. Polan AK: Epidemiologic evidence for transmission of Hodgkin‘s disease. N Engl J Med I(l:499—502. I973. Vianna NJ. Polan AK. Keogh MD et al: Hodgkin‘s dis- ease mortality among physicians. Lancet ii:l3I—I33. I974. Walrath J. Fraumeni JF Jr: Mortality among embalmers. Int J Cancer 3|:407—4I I. I983. Waterhouse J. Muir C. Correa P et al (eds): Cancer Inci— dence in Five Continents. Vol. 3. [ARC Scientific Pub] No. [5. Lyon. France: International Agency for Research on Cancer. I976. Walerhouse J. Muir C. Powell J. et al (edsI: Cancer Inci- dence in Five Continents. Vol. IV. [ARC Scientific Publ No. 42. Lyon. France: International Agency for Research on Cancer. I982. Welton JC. Marr JS. Friedman SM: Association be- tween hepatohiliary cancer and typhoid carrier state. Lancet 1:791—794. I979. Weiden PL. Lerner KG. Gerdes A et al: I’ancytopenia and leukemia in Hodgkin's disease: Report of three cases. Blood 43:571. 1973. Weiss NS: ()vary. In Cancer Epidemiology and Preven- tion (Schottent‘eld D and Fraumeni JF Jr. eds). Phila- delphia: W.B. Saunders. 1982. pp. 871—880. Weiss NS. Sayvetl. TA: Incidence ot‘endometrial cancer in relation to the use of oral contraceptives. N Engl J Med 302:55I—554. I980. Winn DM. Blot WH. Shy CM et al: SnuII~ dipping and oral cancer among women in the southern United States. N Engl J Med 304:745—749. I981. Winn DM. Blot WJ. Shy CM. Fraumeni JF Jr: Occupa- tion and oral cancer among women in the South. Am J Ind Med 3:]61—167. I982. Wittenmore AS. McMillan A: Lung Cancer Among U.S. Uranium Miners: A Reappraisal. J Natl Cancer Inst 71:489—499. I983. Wynder EL. Cornfield J. Schrofi' PD et al: A study of environmental factors in carcinoma of the cervix. Am J Obstet Gynecol 68:l0l6—|047. I954. Wynder EL. HoITmann D. Tobacco. In Cancer Epidc» miology and Prevention tSchottenI‘cld D and Frau- meni JF Jr. eds.). Philadelphia: W.B. Saunders. I982. pp 277—292. Wynder EL. Hultherg S. Jacobsson F et al: Environ- mental factors in cancer of the upper alimentary tract. A Swedish study with special reference to Plummer—Vinson (Paterson-Kelly) syndrome. Cancer 10:470—487. I957. Wynder EL. Mahuchi K. Whilmore WF Jr: Epidemiol» ogy of adenocarcinoma of the kidney. J Natl Cancer Inst 53:]6I9—1634. I974. Wynder EL. Mabuchi K. Whilmore WF: Epidemiology of cancer ofthe prostate. Cancer 28:344-360. I97I. Young JL Jr. Heise HW. Silvcrberg E. Myers MH: Can- cer Incidence. Survival and Mortality for Children Under 15 Years of Age. American Cancer Society. Professional Education Publication. I978. Young JI. Jr. Miller RW: Incidence of malignant tumors in U.S. children. J Pediatr 86:254—258. I975. Young JL Jr. Percy CL. Asire AJ. eds: Surveillance. Epidemiology. and End Results: Incidence and Mor- tality Data. I973—77. DHHS Pub] No. (NIH) III-Z330. Natl Cancer Inst Monogr 57. Washington. DC: Govt I’rint ()fI‘. I98]. Young RC. Knapp RC. Perez. CA: Cancer ot‘the ovary. In Cancer. Principles and Practice of Oncology tI)e~ Vita VT. Hillman S. Rosenberg SA. eds). Phila- delphia: W.B. Saunders. I982. pp. 884—9I3. Zeil HK. Finkle WD: Increased risk ofendometrial ear- cinoma among users of conjugated estrogens. N Engl J Med 293:”64—1167. I975. Ziegler RG. Morris LE. Blot WJ. Pottern LM. Hoover R. Fraumeni JF Jr: Esophageal cancer among black men in Washington. DC. II. Role of nutrition. J Natl Cancer Inst 67:I [99—1206. I9XI. GENERAL UBBAflY-ILC.BERKELEY 0000528213 US DEPARTMENT OF HEALTH AND NIH Publication No. 85-691 HUMAN SERVICES April 1985 Public Health Service National Institutes of Health ,,,,,, ____——¥