Increasing incidence of childhood tumours of the central nervous system in Denmark, 1980-1996

Abstract

The registered incidence rate of childhood central nervous system (CNS) tumours has increased in several countries. It is uncertain whether these increases are biologically real or owing to improved diagnostic methods. We explored the medical records of 626 CNS tumours diagnosed in Danish children between 1980 and 1996. Population-based registers were used to extract data on mortality and background population. Temporal patterns were analysed by regression techniques. Most tumours were verified by computed tomography (78%) or magnetic resonance imaging (14%). Overall, the incidence rate increased by 2.9% per year (95% confidence interval (CI): 1.3;4.5) and the mortality rate increased by 1.4% per year (95% CI: -0.4;3.3). Among children aged 0-4 years, the survival rate after diagnosis remained almost unchanged, whereas among children aged 5-14 years, the 10-year survival rate improved from 59 to 74%. These data suggest that the incidence rate of CNS tumours among Danish children has truly increased, although alternative explanations cannot be excluded.

Figure 1

Annual incidence rates (per million) of CNS tumours among Danish children, 1980–1996, age standardised to the Danish childhood population in 1988. A smoothing spline indicates the temporal pattern.

Figure 2

Percentage of childhood CNS tumours diagnosed in Denmark, 1980–1996, for which CT scanning (dots) or MR scanning (triangles), respectively, was the first method used to verify the tumour. Ten tumours were excluded as they had been verified by both CT and MRI on the same day.

Figure 3

Kaplan–Meier plots of survival after a diagnosis of a CNS tumour in Denmark in 1980–1987 and 1988–1996. The upper panel shows children aged 0–4 years, and the lower panel represents children aged 5–14 years at the time of diagnosis.