Non-Hodgkin's lymphoma : a search for causes

Sammanfattning: NON-HODGKIN'S LYMPHOMA; a search for causes Johanna Adami, MD,MPH Department of Medical Epidemiology, Karolinska Institutet, 17177 Stockholm and Department of Oncology, University Hospital,751 85 Uppsala, Sweden. The aim of this thesis was to investigate some factors of possible importancefor the development of non-Hodgkin's Iymphoma (NHL), namely ultraviolet (UV) light,maternal and perinatal characteristics, blood transfusions, and smoking. To test the UV light hypothesis, a cohort study of secondary occurrence of skincancer (malignant melanoma and squamous cell skin cancer) following NHL, and of NHLfollowing skin cancer, was executed. All incident cancers diagnosed between 1943and 1989 were identified in the nationwide Swedish and Danish cancer registries.Compared to the risks in the general population, the overall risks for squamous cellskin cancer and malignant melanoma subsequent to NHL were significantly elevated5-9-fold and 2-3-fold repectively. Conversely, the risk for NHL was increased 2-foldfollowing squamous cell skin cancer and increased 40% following malignant melanoma. In a population-based nested case-control study based on 1.7 million births recordedin the nationwide Swedish Birth Registry in 1973 through 1989, record-linkage tothe Swedish Cancer Registry disclosed 168 incident childhood NHL cases, for whom839 individually matched controls were randomly selected. Cesarean section, paracervicalblockade and infant endocrine-metabolic disorders were nonsignificantly associatedwith increased risks of developing NHL before the age of 15 years. Maternal hypertensivediseases were linked with a reduced risk for NHL, although not statistically significantly. In a population-based case-control study, we evaluated the role of blood transfusionswithin a cohort of 96,795 patients registered in the Swedish Inpatient Registry between1970 and 1983. Follow-up through 1991 in the Cancer Registry revealed 361 incidentcases of NHL, for whom 705 individually matched controls were randomly selected.Exposure was assessed through linkage to Transfusion Registries at the respectivehospitals. No association was found between blood transfusion and the risk of NHL. To assess the role of smoking in the development of hematolymphoproliferativemalignancies, a cohort study including 333,288 men recorded in the Swedish ConstructionWorkers Registry in 1971 through 1991 was designed. Among these, a total of 1,322incident Iymphoid and hematopoietic neoplasms occurred according to the Swedish CancerRegistry. NHL and multiple myeloma showed no association with tobacco use. The riskof developing Hodgkin's disease was slightly elevated, although there was no clearindication of trends with duration. We used the Cancer-Environment Registry III to analyze latitude, a surrogate measureof UV-light exposure, as a determinant of the risk for NHL. Among 10,381 incidentcases of NHL, we found an increased risk of NHL associated with a more southern locationof the home or workplace. Key words: non-Hodgkin's Iymphoma, malignant melanoma, squamous cell skin cancer,case control study, cohort study, ultraviolet light, maternal factors, perinatalfactors, tobacco use, blood transfusion, occupation. ISBN 91-628-2676-X