New evidence on germ-cell testicular cancer aetiology

Abstract

Testicular cancer has been increasing in incidence for at least 50 years in many populations, but its etiology remains elusive. We investigated several prenatal and postnatal factors in association with germ-cell testicular cancer risk. Moreover, we evaluated recent trends in testicular cancer incidence in Northern European countries.

Using data from birth records and from the Swedish Cancer Register, we carried out a casecontrol study of 628 cases and 2,309 controls. We aimed at investigating the association between perinatal characteristics and testicular cancer, and assessing potential etiological heterogeneity between seminomas and nonseminomas, the two major histological groups of testicular cancer. Gestational duration was inversely associated with testicular cancer risk, whereas men with both high and low birth weight had an excess risk, indicating that intrauterine environment affects the risk of testicular cancer. Seminomas and nonseminomas seemed to have similar risk patterns.

Data from the first study were linked to the Swedish Military Service Conscription Register, which contains information on a medical examination that is mandatory in Sweden for the purpose of military service. The linkage permitted us to get information on body size at two different points in life on 371 cases and 1,238 controls. We found that height at eighteen years old is directly associated with testicular cancer risk. The association persisted after adjusting for perinatal characteristics, suggesting that both foetal life and later periods in life, such as childhood and adolescence, are important time windows for determining lifetime risk for testicular cancer.

A case-control study, including 3,051 cases and 9,007 controls born in Sweden after 1940, was carried out using data from the Multi-Generation Register and the Swedish Cancer Register. We obtained information on number and gender of first-degree relatives of all study subjects. We found that both low birth order and having few siblings is associated with an increased risk of testicular cancer. Since sibship size is correlated with birth order, we performed stratified analyses to disentangle between the effects of the two variables, and found that sibship size is a more important factor. We interpret that these findings are explained by an association between parental fertility and risk of testicular cancer in the offspring.

Data from the Multi -Generation Register and the Cancer Register were also used to investigate the fertility status before and after diagnosis of testicular cancer. Fecundity and the likelihood of fathering dizygotic twins, which is decreased among subfertile subjects, were used as independent measures of fertility of 4,592 cases and 12,154 controls, born in Sweden in 1916 onwards. Prior to diagnosis cases had a decreased number of children, with a lower frequency of dizygotic twinning, indicating that testicular cancer patients have an increased frequency of fertility problems before diagnosis. After diagnosis cases fathered twins more often than controls, probably reflecting an increased use of assisted reproduction techniques.

Finally, the occurrence patterns of testicular cancer in eight Northern European countries were evaluated using data from national Cancer Registries, We found that the incidence of seminomas and nonseminomas is still increasing in all countries analyzed, with the possible exception of Denmark. Moreover, we found that, in Scandinavian countries, the increasing trend is a birth cohort phenomenon also in recent cohorts.