HLA Antigens and Adult T-Cell Leukemia Virus Infection: A Community-Based Study in the Goto Islands, Japan

In order to study the genetical susceptibility to infectionby adult T-cell leukemia virus (ATLV), which is considered tobe a main causative agent of adult T-cell leukemia (ATL), theHLA profile of persons with antibody against ATLV-associatedantigen (anti-ATLA) was compared with that of antibody-negativepersons in villages A and B of the Goto Islands, a typical ATLendemic area in Kyushu, Japan. The frequency (29%) of HLA-B7in persons with anti-ATLA was significantly higher than that(8%) in anti-ATLA negatives (corrected X²=7.73, p<0.01) inone village while the frequency of HLA-B7 was around 10% inboth positive and negative persons in the other village. Theseresults do not support or deny the possibility of the existenceof individuals who are genetically susceptible to ATLV infectionin ATL endemic areas of Japan.     

A Descriptive Epidemiological Study of Hematopoietic Neoplasms in Japan

In 1982, the number of deaths from hematopoietic neoplasms was5,885 for males and 4,237 for females, which corresponded toabout 6%; of all malignant neoplasms. The increase in the age-adjusteddeath rate in the last 10 years was highest for lymphatic leukemia(1.8 times), followed by multiple myeloma (1.7 times), malignantlymphomas (1.3 times) and myelogenic leukemia (1.1 times). Inthe old-age group, i.e., over 70, the death rate for all typesof hematopoetic neoplasms increased markedly during this period,particularly for multiple myeloma. In children, however, anincrease in lymphatic leukemia and a decrease in myelogenicleukemia were observed. Geographical variation showed a markedexcess in the death rates for malignant lymphomas and lymphaticleukemia in the Kyushu district, especially in the middle- andold-age groups. This may be attributed to the high incidenceof adult T-cell leukemia/lymphoma in the southwestern part ofJapan. The age-adjusted death rate for malignant lymphomas wasslightly higher in rural areas than in urban areas in malesin Kyushu.     

Second Malignancy in Stomach Cancer Patients and Its Possible Risk Factors

To identify possible risk factors of multiple primary cancersin stomach cancer patients, a case-control analysis based ondata from medical records was conducted on 73 patients (54 malesand 19 females) with stomach cancer who were affected by oneor more other primary cancers subsequently or concurrently (cases)and 146 patients with stomach cancer alone (controls), by matchingsex, age, time of the operation for stomach cancer and survivalperiod. The relative risks (RR) for cancer history in the father(RR = 3.21), for family history of stomach cancer (RR = 2.52)and for history of hemorrhoids (RR = 3.15) were significantlyincreased in both sexes. The proportion of professionals inmales was significantly higher in cases than in controls. Ina conditional multiple logistic regression analysis for males,the adjusted relative risk (aRR) for family history of cancerin two or more members (aRR = 3.54) was statistically significant.Smoking tended to increase the risk of multiple primary cancersin males, especially in the cases with smoking-related secondmalignancy (RR =4.83). These data suggest the possibility thatboth genetic and environmental factors are involved in the multiplicityof cancer in stomach cancer patients.     

Latitude and Pancreatic Cancer

With regard to the geographical distribution of pancreatic cancermortality, the relationship of the latitude and of the averagetemperature to pancreatic cancer mortality was examined. Bothinternationally and within Japan, a strong positive associationbetween latitude and pancreatic cancer mortality and a strongnegative association with the average temperature were observed.The simple correlation coefficient between pancreatic cancermortality and the latitude was 0.724 for males and 0.725 forfemales, and that between pancreatic cancer mortality and theaverage temperature was –0.773 for males and –0.729for females in 26 countries. In the 47 prefectures in Japan,the coefficient was 0.612, 0.615, –0.589, and –0.630respectively. Most of these relationships remained statisticallysignificant after controlling for per capita consumption offood which was also related to the pancreatic cancer mortalityinternationally, such as sugar, eggs, milk and dairy products,oils and fats, pulse, coffee, and fat. Within Japan, an urban-ruraleffect on the pancreatic cancer mortality was not observed inspite of a higher intake of fat in large cities. There remainsthe possibility that factors related to latitude or averagetemperature other than diet may be involved in the occurrenceof pancreatic cancer.