Etiological insights into the testicular cancer epidemic

Abstract

The aim of this thesis was to search for causes of the testicular cancer epidemic through epidemiologic studies of testicular cancer and possibly related conditions, such as urogenital birth defects and subfertility.

First, we assessed changes in fertility over time by studying time-to-pregnancy among primiparous women in Sweden from 1983 through 1993 using data from the Medical Birth Registry. The analyses revealed a substantial decrease of subfertility over successive maternal birth cohorts. Thus, there was no detectable increment in subfertility problems paralleling the increase in testicular cancer during the study period.

Second, we performed a nested case-control study to evaluate specific perinatal characteristics as risk factors for testicular cancer. Perinatal information about 232 case patients and 904 control subjects was obtained from birth records. We found elevated risks associated with neonatal jaundice and high or low birth weight. When the two histopathologic types of testicular cancer, seminoma and nonseminoma, were analyzed separately, high socioeconomic status, neonatal jaundice, and low birth weight were associated with nonseminomas, whereas high placental weight appeared to increase the risk for seminomas.

Third, we investigated risk factor patterns for cryptorchidism and hypospadias in two case-control studies using record linkage between the Swedish Inpatient and Birth Registries. Cases were 2,782 and 1,220 boys operated for cryptorchidism or hypospadias, respectively. Partly similar risk patterns were revealed for the two conditions: both were positively associated with other congenital malformations and inversely with maternal parity. There was also evidence of a strong joint effect of intrauterine growth retardation and short duration of gestation, with highly increased risks among preterm boys born small-for-gestational-age.

Fourth, seroreactivity against cytomegalovirus (CMV) and Epstein-Barr virus (EBV) was evaluated in relation to testicular cancer risk in a case-control study, nested in a cohort of prospectively collected serum specimens from 293,692 individuals. For 81 cases of testicular cancer and 242 controls, serum IgG titers were determined using enzyme-linked immunosorbent assays and immunofluorescence methods. No association was found between CMV positivity and testicular cancer. For EBV, the risk of testicular cancer was increased almost threefold among individuals seropositive for viral capsid antigen. Power limitations, however, hamper conclusive inference.

Fifth, we conducted a cohort study using anthropometrical data of 477,248 men who attended a tuberculosis-screening program in Norway between 1963 and 1975. Five hundred and fifty three cases of testicular cancer were diagnosed up to 1989. High body mass index (BMI) was associated with a decreased risk of testicular cancer whereas height was positively associated with risk. The associations appeared more pronounced for seminomas than for nonseminomas.

Thus, hormonal exposures may have different effects on different histologic subgroups, and if so, such exposures cannot explain the dramatic rise in testicular cancer incidence, which comprises both types. It is furthermore hypothesized that impaired placental function is important in the etiology of testicular cancer, and that improved survival of growth-retarded and prematurely born neonates may have contributed to the increasing trend.