The present article describes the series of incident primary ovarian tumors in the Life Span Study (LSS) cohort of the Radiation Effects Research Foundation, with particular emphasis on case ascertainment and characterization of histological features of the tumors. We identified 723 ovarian tumors (260 malignant, 463 benign) in 648 individuals of about 70,000 female LSS subjects; 71 cases had more than one ovarian tumor. We histologically confirmed 601 tumors (182 malignant, 419 benign tumors). The most frequent histological type was common epithelial tumor (90.7% for malignant and 59.7% for benign tumors). The distributions of ovarian tumors by histological type were similar to those from other studies. Among malignancies, the frequency of common epithelial types relative to other tumor types increased with radiation dose (p = 0.02). Among benign tumors, the relative frequency of sex-cord stromal tumors increased with radiation dose (p = 0.04). The women with mucinous cancer had better survival than those with serous cancers (p = 0.03). Within tumor types, there was no consistent pattern of survival by radiation dose. Variations in histological types of ovarian tumors in response to radiation dose, suggested by the case series data need to be followed up by population-based incidence analysis.
SP-29 The incidence rates of adult T-cell leukemia/lymphoma (ATL) and non-Hodgkin lymphoma (NHL) were evaluated in Kamigoto Islands, a human T-cell lymphotropic virus type-I (HTLV-I) endemic area in Nagasaki, Japan. The total population of the study area in 1990 included 12,820 men and 14,050 women. Between January 1985 and August 1996, sera from 8,771 men and 9,714 women living in the study area were screened for antibodies to HTLV-I, using a particle agglutination assay. The overall HTLV-I seroprevalence was 14.8% for men and 18.6% for women. The person-years at risk of HTLV-I carriers were estimated on the basis of the total population and the town-, sex- and 5-year-age-specific prevalence of HTLV-I seropositivity. Using a population-based cancer registry, we identified 37 cases of ATL and 28 cases of other NHL who were diagnosed during 1985-1994. The crude annual incidence rate of ATL among 100,000 HTLV-I carriers aged 30 or older was estimated at 138.0 (95% confidence interval [CI] 87.5-207.1) for men and 52.7 (95% CI 28.8-88.4) for women, with a significant sex difference even after adjustment for age (p=0.002). The lifetime risk of the disease, i.e., the cumulative risk from 30 to 79 years of age, was estimated at approximately 6.5% (95% CI 3.7-9.2) for men and 1.9% (95% CI 0.8-2.9) for women. ATL accounted for approximately 55-61% of the total NHL incidence in the study area.
In medical care systems for cancer, it is important to consider the issues of standardization and centralization. In this study, we employed the Nagasaki Cancer Registry, which has a high registry rate, to investigate standardization and centralization for five major cancers, in addition to childhood malignancies (which are often rare types). Subjects were patients diagnosed with cancer and registered in the Nagasaki Cancer Registry between 1985 and 2004. For standardization, we calculated a Preference Index and five-year survival rate, and for centralization we investigated Pareto curves and Gini coefficients as well as the annual average number of cases per hospital. Results suggested that patients migrate to medical service areas different from where they reside in order to receive treatment at facilities thought to have a better record of treatment. In addition, while the number of patients and treatment facilities for childhood cancer was decreasing due to a decline in the number of children, the centralized tendency differed for the 12 diagnoses assessed. By conducting analyses based on population-based cancer registries using the evaluation methods employed in this study, it should be possible to investigate patients' migrant patterns, as well as to develop systems for providing medical care in secondary medical service areas.
P-798 Introduction: The purpose of this study was to clarify whether lifestyle may modify the effects of ionizing radiation exposure on the risk of lung cancer. Numerous laboratory and animal studies have described the protective effects of dietary factors supplemented before radiation exposure, examining their radio-protective effects. However, human studies are very limited and little is known about the lifestyle effects after radiation exposure on radiation damage and cancer risks. Methods: A cohort of 38,430 atomic-bomb survivors of Hiroshima and Nagasaki, for whom radiation dose estimates were currently available, had their lifestyle (smoking, drinking, diet) assessed in 1980. They were followed during 20 years for incidence of lung cancer. The associations between lifestyle factors and cancer risk were examined using a Poisson regression analysis. The combined effect of significant lifestyle factors and radiation exposure on cancer risk was examined in additive and multiplicative models. Results: During the 505,849 person-years of follow-up, 702 cases of primary lung cancer were identified. When examining each lifestyle factor individually, lung cancer risks increased significantly in current tobacco smokers and past smokers as compared to non-smokers (current smokers: Relative Risk RR (95% CI) 4.02 (3.21–5.02); past smokers: RR: 1.56 (1.15–2.13)). Lung cancer risks decreased significantly in those with a daily intake of fruit, as compared to once a week intake (RR: 0.79 (0.65–0.96)). The risks decreased with daily vegetables intake, but not at the significant level (RR: 0.86 (0.70–1.06)). Lung cancer risk was not associated with alcohol or green tea consumption. The observed effects of smoking and radiation were greater than the anticipated effects, suggesting a synergism or a positive interaction. While the observed effects of fruit consumption and radiation were lower than the anticipated effects, suggesting an antagonism or a negative interaction. There was no significant evidence for rejecting either the additive or the multiplicative model. Discussion and Conclusion: Those exposed to radiation and who never smoked tobacco presented a lower lung cancer risk than the ever-smokers exposed to radiation. This finding may be explained by the fact that the genotoxic substances of tobacco smoke may enhance the carcinogenic effect of ionizing radiation. On the other hand, a low fruit intake was not detrimental among those exposed to radiation. In conclusion, in this large cohort of atomic-bomb survivors, no smoking benefited those exposed to ionizing radiation in reducing their risk of lung cancer.TABLE 1: Combined Effects of Smoking or Fruit Intake and Radiation Exposure on Lung Cancer Risk
