Using data from the Japan Children's Cancer Registry, we estimated the age-specific incidence rates of neuroblastoma. Before the neuroblastoma screening program started in 1985, the age-standardized incidence rates of neuroblastoma ranged from 7.5 to 9. I x for children under I5 years of age. After the introduction of the screening program, the annual incidence rate rose to 19.5 X The annual incidence rate for neuroblastoma in children under I year of age was 150.60 X I OW6 in the years I989 to 1992. whereas the incidence rates only varied between 23.6 and 34. I 3 X I 0-6 in the 3 preceding 5-year periods. This increase in incidence for infants was accompanied by a minor decrease in incidence for children 2 to 3 years of age.However, this decrease may only partly explain the large increase in incidence for infants. Therefore, we suggest that screening may result in the detection of otherwise spontaneously regressing tumors. It is urgent to determine the contribution of screening to decreasing mortality before deciding whether this screening program should be continued.o I994 Wiley-Liss, Inc.Since its inception, there have been controversies regarding the effectiveness of neuroblastoma screening for 6-month-old infants (Woods and Tuchman, 1987; Mauer, 1988;Tuchman et al., 1990;Murphy et al., 1991). Several reports have shown that the relative frequencies of the tumor failed to decrease in older children following screening (Bessho et al., 1991) or have suggested that a large number of the tumors found by screening might otherwise have spontaneously regressed (Nishi et al., 1995). We have estimated the age-specific incidence of neuroblastoma, using data from the Japan Children's Cancer Registry, to determine whether screening has resulted in any changes in the age-specific annual incidence rate.
MATERIAL AND METHODSNumbers of neuroblastoma cases registered with the Japan Children's Cancer Registry (JCCR) between 1969 and 1992 were obtained from 3 volumes of 5-year reports (JCCR Committee, 1975, 1983 and from annual reports of the JCCR (JCCR Committee, 1987. Neuroblastoma, ganglioneuroblastoma and ganglioneuroma were collectively included as "neuroblastoma". Ganglioneuroma was included because neuroblastoma and ganglioneuroblastoma could mature to ganglioneuroma in vivo when they were left untreated (Haas et al., 1988).Sizes of the population in each age group were calculated by year from vital statistics published by the Statistics and Information Department, Minister's Secretariat, Ministry of Health and Welfare of Japan. The population of infants under 1 year of age was the number of births. The population of a particular age group in a particular year was calculated by subtracting the number of deaths from the population data of the I-year-younger age group in the previous year.The results of screening are reported annually by the Neuroblastoma Screening Committee.
Estimation of registration rate of JCCRThe JCCR is not population-based and its exact registration rate is not known. Therefore, an estimation of ...