Household water source and the risk of childhood brain tumours: results of the SEARCH International Brain Tumor Study.

BACKGROUND: The period in utero is a time of increased vulnerability. Offspring of pregnant women exposed to carcinogenic substances in drinking water may be more likely to develop cancer. We examined whether household water source and the presence of nitrates or nitrites in residential water were associated with increased risks of childhood brain tumours (CBT). METHODS: We used data from a multicentre, case-control study with maternal information on residential water source, and nitrate/nitrite levels of tap water measured by dipstick. Subjects included 836 CBT cases and 1485 controls from five countries. RESULTS: The risks of CBT associated with reliance on well water (versus public water) during pregnancy varied widely, with significantly increased risks noted in two (of seven) regions and a decreased risk observed in one region. CBT risk did not increase with increasing nitrate levels. However, our results based on tap water tested in the pregnancy residences suggest the risk of astrocytoma may be associated with increasing levels of nitrite (odds ratio [OR] = 4.3, 95% CI: 1.4, 12.6 for nitrite levels of 1-<5 mg/l nitrite ion; OR = 5.7, 95% CI: 1.2, 27.2 of nitrite > or =5 mg/l). CONCLUSIONS: These results should be interpreted with caution because women's recollection of water sources may have contained inaccuracies, and nitrate and nitrite measurements, available for only a portion of subjects, were often obtained years after the pregnancies occurred. However, our results suggest a need for closer evaluation of well water content in some regions and the possibility that a nitrite-related water exposure may be associated with CBT.

Relation of childhood brain tumors to exposure of parents and children to tobacco smoke: the SEARCH international case-control study. Surveillance of Environmental Aspects Related to Cancer in Humans.

The etiology of childhood brain tumors (CBTs) remains unknown. Tobacco smoke contains several known carcinogens and can induce DNA adducts in human placenta and hemoglobin adducts in fetuses. We present the results of an international case-control study to evaluate the association between CBTs and exposure of parents and children to cigarette smoke. The study was undertaken as part of the SEARCH program of the IARC. Nine centers in 7 countries were involved. The studies mainly covered the 1980s and early 1990s. Cases (1,218, ages 0-19 years) were children newly diagnosed with a primary brain tumor; there were 2,223 population-based controls. Most mothers who agreed to participate were interviewed in person at home. Odds ratios (ORs) were calculated by unconditional logistic regression, adjusted for age, sex and center, for all types of CBT combined, 4 CBT histotypes, 5 age groups and each center. There was no association between the risk of brain tumors in the child and parental smoking prior to pregnancy, maternal smoking or regular exposure to others' cigarette smoke during pregnancy at home or at work, or passive smoking by the child during the first year of life. These results did not change considering the child's age at diagnosis, the histologic type of tumor or center.

Non-Hodgkin's lymphoma and specific pesticide exposures in men: cross-Canada study of pesticides and health.

Our objective in the study was to investigate the putative associations of specific pesticides with non-Hodgkin's Lymphoma [NHL; International Classification of Diseases, version 9 (ICD-9) 200, 202]. We conducted a Canadian multicenter population-based incident, case (n = 517)-control (n = 1506) study among men in a diversity of occupations using an initial postal questionnaire followed by a telephone interview for those reporting pesticide exposure of 10 h/year or more, and a 15% random sample of the remainder. Adjusted odds ratios (ORs) were computed using conditional logistic regression stratified by the matching variables of age and province of residence, and subsequently adjusted for statistically significant medical variables (history of measles, mumps, cancer, allergy desensitization treatment, and a positive history of cancer in first-degree relatives). We found that among major chemical classes of herbicides, the risk of NHL was statistically significantly increased by exposure to phenoxyherbicides [OR, 1.38; 95% confidence interval (CI), 1.06-1.81] and to dicamba (OR, 1.88; 95% CI, 1.32-2.68). Exposure to carbamate (OR, 1.92; 95% CI, 1.22-3.04) and to organophosphorus insecticides (OR, 1.73; 95% CI, 1.27-2.36), amide fungicides, and the fumigant carbon tetrachloride (OR, 2.42; 95% CI, 1.19-5.14) statistically significantly increased risk. Among individual compounds, in multivariate analyses, the risk of NHL was statistically significantly increased by exposure to the herbicides 2,4-dichlorophenoxyacetic acid (2,4-D; OR, 1.32; 95% CI, 1.01-1.73), mecoprop (OR, 2.33; 95% CI, 1.58-3.44), and dicamba (OR, 1.68; 95% CI, 1.00-2.81); to the insecticides malathion (OR, 1.83; 95% CI, 1.31-2.55), 1,1,1-trichloro-2,2-bis (4-chlorophenyl) ethane (DDT), carbaryl (OR, 2.11; 95% CI, 1.21-3.69), aldrin, and lindane; and to the fungicides captan and sulfur compounds. In additional multivariate models, which included exposure to other major chemical classes or individual pesticides, personal antecedent cancer, a history of cancer among first-degree relatives, and exposure to mixtures containing dicamba (OR, 1.96; 95% CI, 1.40-2.75) or to mecoprop (OR, 2.22; 95% CI, 1.49-3.29) and to aldrin (OR, 3.42; 95% CI, 1.18-9.95) were significant independent predictors of an increased risk for NHL, whereas a personal history of measles and of allergy desensitization treatments lowered the risk. We concluded that NHL was associated with specific pesticides after adjustment for other independent predictors.

Parental occupations and childhood brain tumors: results of an international case-control study.

OBJECTIVE: To evaluate the role of parental occupations in the etiology of childhood brain tumors (CBT). METHODS: Population-based case-control studies were conducted concurrently in seven countries under the coordination of the International Agency for Research on Cancer, gathering 1,218 cases and 2,223 controls. We report here the findings related to parental occupations during the 5-year period before the child's birth. Risk estimates related to a number of paternal and maternal occupations were obtained by unconditional logistic regression adjusted for age, sex, year of birth, and center, for all types of CBT combined and for the subgroups of astroglial, primitive neuroectodermal tumors (PNET), and other glial tumors. RESULTS: An increased risk in relation with agricultural work was seen for all CBT combined and for other glial tumors. Increased risks for all tumors and PNET were seen for paternal occupation as an electrician; the same pattern held for maternal occupation when children under 5 were selected. Paternal occupation as a driver or mechanic, and maternal work in an environment related to motor-vehicles were associated with an increased risk for all CBT and astroglial tumors. More case mothers compared to control mothers were employed in the textile industry. CONCLUSION: Our study reinforces previous findings relative to the role of parental work in agriculture, electricity, or motor-vehicle related occupations and maternal work in the textile industry. It does not confirm previous associations with work environments including aerospace, the chemical industry, or the food industry, or with maternal occupation as a hairdresser, a nurse, or a sewing machinist, and paternal occupation as a welder.

SEARCH international case-control study of childhood brain tumours: role of index pregnancy and birth, and mother's reproductive history.

A series of co-ordinated population-based case-control studies of childhood brain tumours (CBT) was undertaken under the auspices of the Surveillance of Environmental Aspects Related to Cancer in Humans (SEARCH) programme of the International Agency for Research on Cancer (IARC) to evaluate, inter alia, the risk in relation to characteristics of the index pregnancy and birth, and maternal reproductive history. Subjects comprised 1218 cases aged 0-19 years and 2223 controls. Risk estimates were calculated by unconditional logistic regression, adjusted for age, sex, centre and mother's years of schooling, for all types of CBT combined as well as for four groups defined by histopathology (astrologlial tumours, primitive neuroectodermal tumours of the brain, 'other glial' tumours and 'other histological types') and for five age groups (0-1, 0-4, 5-9, 10-14, 15-19 years). Use of anaesthetic 'gas' was associated with an increased risk of CBT (OR = 1.5, 95% CI [1.1, 2.0]), apparent in children aged 0-4 years (OR = 2.4, 95% CI [1.4, 4.1]) and for astroglial tumours (OR = 1.6, 95% CI [1.1, 2.2]) with non-significantly increased relative risks for each of the other histological groups. However, not all centre-specific relative risks were elevated. No other aspect of the index pregnancy, delivery and early neonatal period or of the mother's previous reproductive history was associated with risk for CBT.

An international case-control study of adult glioma and meningioma: the role of head trauma.

BACKGROUND: Increased brain tumour risk after head trauma suggested by case reports and clinical series has been previously studied epidemiologically with mixed results. An international multicentre case-control study investigated the role of head trauma from injury or sports participation in adult brain tumour risk. METHODS: In all, 1178 glioma and 330 meningioma cases were individually or frequency matched to 2236 controls. Only exposures that occurred at least 5 years before diagnosis and head injuries that received medical attention were considered. RESULTS: Risk for ever having experienced a head injury was highest for male meningiomas (odds ratio [OR] = 1.5, 95% confidence interval [CI] : 0.9-2.6) but was lower for 'serious' injuries, i.e. those causing loss of consciousness, loss of memory or hospitalization (OR = 1.2, 95% CI: 0.6-2.3). Among male meningiomas, latency of 15 to 24 years significantly increased risk (OR = 5.4, 95% CI: 1.7-16.6), and risk was elevated among those who participated in sports most correlated with head injury (OR = 1.9, 95% CI: 0.7-5.3). Odds ratios were lower for male gliomas (OR = 1.2, 95% CI : 0.9-1.5 for any injury; OR = 1.1, 95% CI: 0.7-1.6 for serious injuries) and in females in general. CONCLUSIONS: Evidence for elevated brain tumour risk after head trauma was strongest for meningiomas in men. Findings related to sports should be interpreted cautiously due to cultural variability in our data and our lack of complete data on physical exercise in general which appeared to be protective.

Combination chemotherapy of oral 5'-deoxy-5-fluorouridine and cisplatin in advanced gastric cancer: a phase II study.

This study was designed to test the activity and feasibility of 5'-deoxy-5-fluorouridine (5'-DFUR) and cisplatin combination therapy in the treatment of advanced gastric cancer. Nineteen patients with inoperable and/or metastatic gastric cancer, which was histologically proven, were orally administered 5'-DFUR 1,200 mg/m2 on days 1 to 4 and days 15-18 combined with 70 mg/m2 of cisplatin being repeated every 4 weeks. Five partial responses (PRs) were achieved. Seven patients had stable disease and 6 progressed on therapy. The overall response rate was 27.7% (95% confidence interval: 9.69% to 53.5%). The median survival duration of all 18 patients was 25 weeks (9-64). The majority of patients had WHO grade I/II toxicity, but there was no treatment-related death. These data support that the combinations of oral 5'-DFUR and cisplatin are well tolerable and have a moderate activity with low toxicity in the treatment of advanced gastric cancer.

Case-control study of residential radon and lung cancer in Winnipeg, Manitoba, Canada.

A case-control study of lung cancer in relation to exposure to radon in homes in Winnipeg, Manitoba, Canada, was conducted during 1983-1990. In total, 738 individuals with histologically confirmed incident cases of lung cancer were interviewed, along with 738 controls matched on age (+/- 5 years) and sex. Radon dosimeters were placed in all residences in which the study subjects had reported living within the Winnipeg metropolitan area for at least 1 year. Radon dosimetry was done by means of integrated alpha-track measurements over a 1-year period. In the homes monitored, the average level of radon-222 was about 120 becquerels (Bq)/m3 in the bedroom area and 200 Bq/m3 in the basement. After adjusting for cigarette smoking and education, no increase in the relative risk for any of the histologic types of lung cancer observed among the cases was detected in relation to cumulative exposure to radon.

Radiation dose and breast cancer risk in patients treated for cancer of the cervix.

The relationship between breast cancer and radiation treatment for cervical cancer was evaluated in an international study of 953 women who subsequently developed breast cancer and 1,806 matched controls. Radiation doses to the breast (average 0.31 Gy) and ovaries (average 32 Gy) were reconstructed for exposed subjects on the basis of their original radiotherapy records. Overall, 88% of the breast cancer cases and 89% of the controls received radiation treatment [relative risk (RR) = 0.88; 95% confidence interval (CI) = 0.7-1.2]. Among women with intact ovaries (561 cases, 1,037 controls), radiotherapy was linked to a significant 35% reduction in breast cancer risk, attributable in all likelihood to the cessation of ovarian function. Ovarian doses of 6 Gy were sufficient to reduce breast cancer risk but larger doses did not reduce risk further. This saturation-type response is probably due to the killing of a critical number of ovarian cells. Cervical cancer patients without ovaries (145 cases, 284 controls) were analyzed separately because such women are at especially low natural risk for breast cancer development. In theory, any effect of low-dose breast exposure, received incidentally during treatment for cervical cancer, should be more readily detectable. Among women without ovaries, there was a slight increase in breast cancer risk (RR = 1.07; 95% CI = 0.6-2.0), and a suggestion of a dose response with the RR being 1.0, 0.7, 1.5 and 3.1 for breast doses of 0, 0.01-0.24, 0.25-0.49 and 0.50+ Gy, respectively. However, this trend of increasing RR was not statistically significant. If low-dose radiation increases the risk of breast cancer among women over age 40 years, it appears that the risk is much lower than would be predicted from studies of younger women exposed to higher doses.

Radiation dose and second cancer risk in patients treated for cancer of the cervix.

The risk of cancer associated with a broad range of organ doses was estimated in an international study of women with cervical cancer. Among 150,000 patients reported to one of 19 population-based cancer registries or treated in any of 20 oncology clinics, 4188 women with second cancers and 6880 matched controls were selected for detailed study. Radiation doses for selected organs were reconstructed for each patient on the basis of her original radiotherapy records. Very high doses, on the order of several hundred gray, were found to increase the risk of cancers of the bladder [relative risk (RR) = 4.0], rectum (RR = 1.8), vagina (RR = 2.7), and possibly bone (RR = 1.3), uterine corpus (RR = 1.3), cecum (RR = 1.5), and non-Hodgkin's lymphoma (RR = 2.5). For all female genital cancers taken together, a sharp dose-response gradient was observed, reaching fivefold for doses more than 150 Gy. Several gray increased the risk of stomach cancer (RR = 2.1) and leukemia (RR = 2.0). Although cancer of the pancreas was elevated, there was no evidence of a dose-dependent risk. Cancer of the kidney was significantly increased among 15-year survivors. A nonsignificant twofold risk of radiogenic thyroid cancer was observed following an average dose of only 0.11 Gy. Breast cancer was not increased overall, despite an average dose of 0.31 Gy and 953 cases available for evaluation (RR = 0.9); there was, however, a weak suggestion of a dose response among women whose ovaries had been surgically removed. Doses greater than 6 Gy to the ovaries reduced breast cancer risk by 44%. A significant deficit of ovarian cancer was observed within 5 years of radiotherapy; in contrast, a dose response was suggested among 10-year survivors. Radiation was not found to increase the overall risk of cancers of the small intestine, colon, ovary, vulva, connective tissue, breast, Hodgkin's disease, multiple myeloma, or chronic lymphocytic leukemia. For most cancers associated with radiation, risks were highest among long-term survivors and appeared concentrated among women irradiated at relatively younger ages.

Does family history of breast cancer improve survival among patients with breast cancer?

Overall cancer mortality to December 1985 among 291 patients whose breast cancer was diagnosed between 1971 and 1974 and who were interviewed shortly after diagnosis was 39.9% (116 deaths). In this study population a positive maternal family history was strongly associated with breast cancer: The odds ratio for patients versus controls of having a mother with breast cancer was 3.32 (95% confidence limits 1.64 and 6.72); the odds ratio of having a mother, sister, or maternal aunt with breast cancer was 1.92 (95% confidence limits 1.27 and 2.91). However, family history was not associated with stage at diagnosis, which is the most important survival factor (53.6% of patients with a family history and 51.7% without were at a local stage at diagnosis). Survival was better, although not significantly so, among women with maternal relatives with breast cancer. The relative risk of dying of cancer, adjusted for confounding factors, was 1.40 for women without versus with a family history; the difference in survival was not statistically significant.

Radiation dose and leukemia risk in patients treated for cancer of the cervix.

To quantify the risk of radiation-induced leukemia and provide further information on the nature of the relationship between dose and response, a case-control study was undertaken in a cohort of over 150,000 women with invasive cancer of the uterine cervix. The cases either were reported to one of 17 population-based cancer registries or were treated in any of 16 oncologic clinics in Canada, Europe, and the United States. Four controls were individually matched to each of 195 cases of leukemia on the basis of age and calendar year when diagnosed with cervical cancer and survival time. Leukemia diagnoses were verified by one hematologist. Radiation dose to active bone marrow was estimated by medical physicists on the basis of the original radiotherapy records of study subjects. The risk of chronic lymphocytic leukemia, one of the few malignancies without evidence for an association with ionizing radiation, was not increased [relative risk (RR) = 1.03; n = 52]. However, for all other forms of leukemia taken together (n = 143), a twofold risk was evident (RR = 2.0; 90% confidence interval = 1.0-4.2). Risk increased with increasing radiation dose until average doses of about 400 rad (4 Gy) were reached and then decreased at higher doses. This pattern is consistent with experimental data for which the down-turn in risk at high doses has been interpreted as due to killing of potentially leukemic cells. The dose-response information was modeled with various RR functions, accounting for the nonhomogeneous distribution of radiation dose during radiotherapy. The local radiation doses to each of 14 bone marrow compartments for each patient were incorporated in the models, and the corresponding risks were summed. A good fit to the observed data was obtained with a linear-exponential function, which included a positive linear induction term and a negative exponential term. The estimate of the excess RR per rad was 0.9%, and the estimated RR at 100 rad (1 Gy) was 1.7. The model proposed in this study of risk proportional to mass exposed and of risk to an individual given by the sum of incremental risks to anatomic sites appears to be applicable to a wide range of dose distributions. Furthermore, the pattern of leukemia incidence associated with different levels of radiation dose is consistent with a model postulating increasing risk with increasing exposure, modified at high doses by increased frequency of cell death, which reduces risk.

Epidemiology of primary intracranial neoplasms in Manitoba, Canada.

The incidence of primary intracranial tumors in Manitoba, Canada was reviewed. From 1980 through 1985, 657 tumors were diagnosed. The crude incidence rates were 10.2/100,000 for males and 10.8/100,000 for females. The three most common tumors were: astrocytoma 281 (43%), meningioma 145 (22%), and pituitary adenoma 111 (17%). Average annual incidence rates for all tumors showed a bimodal distribution with one peak in the 0-4 age group (4.2/100,000), and the other in the 60-69 age group (27.2/100,000). For malignant astrocytoma, the age-specific annual incidence rate increased to the seventh decade where it reached a peak of 14.3/100,000. The incidence of benign astrocytoma remained relatively constant with age at 1.1/100,000. The annual incidence of meningioma increased with age up to the eighth decade reaching 7.2/100,000. Of the 145 meningiomas, 56 (39%) were meningotheliomatous, 48 (33%) transitional, 10 (7%) malignant, 7 (5%) fibroblastic, 6 (4%) psammomatous, 3 (2%) angioblastic, and 15 (10%) lacked pathologic diagnosis. The annual incidence of pituitary adenoma showed two peaks, the first occurring in the third decade (2.6/100,000) and the second in the eighth decade (3.2/100,000). Although the incidence of meningioma was relatively high, the clinical features and pathologic patterns of these tumors were not unlike those previously reported in the literature.

Second malignancies following testicular cancer, ovarian cancer and Hodgkin's disease: an international collaborative study among cancer registries.

Eleven population-based cancer registries tabulated second cancers among 133,411 patients diagnosed with testicular cancer, ovarian cancer or Hodgkin's disease between 1945 and 1984. Overall, 3,157 second cancers were observed, as compared with 2,420 expected at least one year after the first cancer. Survivors of testicular and ovarian cancer experienced 30% and 20% more cancers respectively than the general population comparison group, and patients previously diagnosed with Hodgkin's disease had an 80% excess of cancer. No information was available either on treatment for the first cancer, or other risk factors. However, temporal patterns in the risk of specific second cancers were analysed, with particular reference to the possible role of therapy for the first cancer. Leukaemia of the acute or non-lymphatic type, which has been previously linked to alkylating agent therapy, occurred in excess following all 3 first cancers, as did non-Hodgkin's lymphoma (overall relative risks of 6.1 and 1.8 respectively, with considerably higher relative risks following Hodgkin's disease). Other cancers for which important and plausibly therapy-induced excesses occurred were lung cancer following Hodgkin's disease (relative risk 1.9), breast cancer following Hodgkin's disease (relative risk 1.4) and bladder cancer following ovarian cancer and Hodgkin's disease (relative risks 1.7 and 2.2 in women, respectively). Rarer sites at which striking excesses occurred were the salivary gland, thyroid, bone and connective tissue. There were smaller, but clear excesses for cancers of the rectum and colon following ovarian cancer and testicular cancer, skin cancer following Hodgkin's disease, and kidney cancer following ovarian cancer. Overdiagnosis, misclassification of metastases and confounding by other risk factors were all considered as explanations of observed excesses. Nonetheless, it appeared that there are clear excess risks for cancers other than acute leukaemia which must be ascribed to therapy for the first cancer, especially in view of the possible under-reporting in registry material. Case-control studies are under way to provide information on the role of specific aspects of therapy.

Consumption of precursors of N-nitroso compounds and human gastric cancer.

It has been hypothesized that dietary nitrate and nitrite are converted in the stomach to nitrous acid, which reacts with secondary amines and amides to form nitrosamines and nitrosamides, compounds frequently demonstrated to be carcinogenic in animals, and that vitamins C and E inhibit N-nitroso product formation by chemically reducing nitrous acid. This hypothesis and others were tested in a case-control study (controls were individually matched by age, sex and area of residence), utilizing a standardized, quantitative, dietary history questionnaire interview. Daily nutrient consumption values were calculated from interview responses, and continuous conditional logistic regression was used for the data analysis. Significant findings are as follows: (1) Average daily consumption of nitrite, chocolate and carbohydrate was associated with increasing trends in risk. (2) While citrus fruit intake appeared to be somewhat protective, any protective effect of vitamin C intake was less apparent, and of vitamin E, not at all apparent. (3) Consumption of dietary fibre was negatively associated with gastric cancer risk. These findings appear to implicate a number of dietary components, including nitrite consumption, in the genesis of gastric cancer in humans.

Dietary factors and the incidence of cancer of the stomach.

A case-control study of diet and stomach cancer was conducted during 1979-1982 in Toronto, Winnipeg, and St. John's Canada. Two hundred forty-six histologically verified cancer cases were individually matched by age, sex, and area of residence to 246 randomly selected population controls. Daily nutrient consumption values were calculated from quantitative diet history questionnaire data through use of the US Department of Agriculture Food Composition Data Bank, which was extended and modified for Canadian items. For the analysis, continuous conditional logistic regression methods were used. It was found that consumption of dietary fiber was associated with decreased risk of gastric cancer; the odds ratio estimate of trend was 0.40/10 g average daily intake of fiber (i.e., 0.40(1.5)/15 g, etc.) (p less than 10(-8)). Also, average daily consumption of nitrite, chocolate, and carbohydrate was associated with increasing trends in risk, with odds ratio estimates, respectively, 2.6/mg (p less than 10(-4)), 1.8/10 g (p less than 10(-4)), and 1.5/100 g (p = 0.015). While citrus fruit intake appeared to be somewhat protective (odds ratio = 0.75/100 g daily average, p = 0.0056), vitamin C intake was less so, and vitamin E not at all. Thus, a number of dietary components seem to be implicated in the pathogenesis of stomach cancer.

Cancer risks among residents of Manitoba Indian reserves, 1970-79.

A descriptive epidemiologic study of malignant neoplasms among residents of Indian reserves in Manitoba from 1970 to 1979 based on the Manitoba Cancer Registry revealed an unusual pattern. There was a greater risk for kidney cancer in both sexes and for gallbladder and invasive cervical cancer in women. The risk was reduced, however, for cancer of the lung in men and of the breast in women, cancers with a high incidence in the general Canadian population. Overall the risk for cancer was lower in both sexes. The results are compared with those of other studies in Indians, and possible exposure to risk factors in this population is discussed.

Second cancers following radiation treatment for cervical cancer. An international collaboration among cancer registries.

The numbers of second cancers among 182,040 women treated for cervical cancer that were reported to 15 cancer registries in 8 countries were compared to the numbers expected had the same risk prevailed as in the general population. A small 9% excess of second cancers (5,146 observed vs. 4,736 expected) occurred 1 or more years after treatment. Large radiation doses experienced by 82,616 women did not dramatically alter their risk of developing a second cancer; at most, about 162 of 3,324 second cancers (approximately equal to 5%) could be attributed to radiation. The relative risk (RR = 1.1) for developing cancer in organs close to the cervix that had received high radiation exposures--most notably, the bladder, rectum, uterine corpus, ovary, small intestine, bone, and connective tissue--and for developing multiple myeloma increased with time since treatment. No similar increase was seen for 99,424 women not treated with radiation. Only a slight excess of acute and non-lymphocytic leukemia was found among irradiated women (RR = 1.3), and substantially fewer cases were observed than expected on the basis of current radiation risk estimates. The small risk of leukemia may be associated with low doses of radiation absorbed by the bone marrow outside the pelvis, inasmuch as the marrow in the pelvis may have been destroyed or rendered inactive by very large radiotherapy exposures. There was little evidence of a radiation effect for cancers of the stomach, colon, liver, and gallbladder, for melanoma and other skin cancers, or for chronic lymphocytic leukemia despite substantial exposures. An excess of thyroid cancer possibly was related to the low dose received by this organ. Ovarian damage caused by radiation may have been responsible for a low breast cancer risk (RR = 0.7), which was evident even among postmenopausal women. A substantial excess of lung cancer (RR = 3.7) largely may be due to misclassification of metastases and the confounding influence of cigarette smoking. Women who were under 30 or over 50 years of age when irradiated were at greatest absolute risk for developing a second cancer. The RR, however, was higher among those under age 30 years at exposure (RR = 3.9) than among older women. The expression period for radiation-induced solid tumors appeared to continue to the end of life.