ABSTRACT

The studies included in this PhD dissertation were carried out at the Department of Epidemiology Research, Statens Serum Institut, and focus on characterising the occurrence of childhood leukaemia, timing of critical genetic events involved in childhood leukaemia development, and the identification of pregnancy-related risk factors for childhood leukaemia with particular emphasis on acute lymphoblastic leukaemia (ALL). Using a unique population-based register of all cases of leukaemia diagnosed in children in the Nordic countries since 1982 (Nordic childhood leukaemia database), we demonstrated stable incidence rates of childhood leukaemia over the last 20 years. This observation suggests that no marked changes in exposures to risk factors for childhood leukaemia have occurred during this period. In a molecular-biological analysis of neonatal blood samples, we found evidence that the development of t(12;21) B-precursor ALL may be initiated in utero. Together with other information this observation is in accordance with the hypothesis that ALL in childhood develops as a multiple step process involving both pre- and postnatal genetic events. In two studies, we observed a statistically significant correlation between birth weight and ALL risk, apparent for all ALL subtypes and diagnostic age groups. Using information on families, we found that children with leukaemia do not weigh more than their siblings at birth, but that siblings of children with leukaemia weigh more at birth than children in general. These findings are consistent with the hypothesis that increasing birth weight modifies the leukaemia risk through proliferative stress and/or by correlating with number of cells at risk of critical genetic aberrations. Our study also showed a reduced risk of ALL with increasing birth order. This association is in accordance with the so-called "delayed infection hypothesis", proposing that postponed exposure to common infections may increase the risk of childhood ALL. Future studies should aim at determining the frequency of the most common genetic traits associated with leukaemia in childhood in healthy newborns, preferably in different populations, and at identifying risk factors for these, possibly initiating, genetic aberrations.