TOPICAL TOPICS

Brain Cancer Incidence in Children:
Time to Look Beyond the Trends 

Medical and Pediatric Oncology 33:110-112 1999 James G. Gurney, PhD'

Division of Pediatric Epidemiology/Clinical Research, Department of Pediatrics. University of Minnesota; and University of Minnesota Cancer Center. Minneapolis, Minnesota

Correspondence to: James G. Gurney, Division of Pediatric Epidemiology, University of Minnesota. 130(1 S. 2nd St., Suite 31)0. Minneapolis, MN 5545-1-1015. E-mail: gurney@epi.umn.edu 

Received 26 January 1999; Accepted 5 March 1999

Primary malignant tumors of the brain, because of the relatively poor prognosis and the substantial morbidity among many who survive, constitute a particularly ominous group of neoplasms. CNS malignancies, of which over 93% are located in the brain, represent 21 % of all malignancies in children younger than 15 years of age in the United States (U.S.). The annual age-adjusted incidence rate is currently about 32 per million children [1, 2]. Nonmalignant intracranial neoplasms, such as craniopharyngiomas, benign meningiomas, pituitary tumors. and choroid plexus papillomas, increase the annual CNS tumor burden among children by an additional 7 per million [2]. Public and scientific concern about rising incidence rates of malignant brain tumors (hereafter called "brain cancer") has been heightened in recent years [3-8]. Indeed, data from the United States [9,10], Europe [11-13], and Australia [14] have shown that brain cancer incidence rates have increased substantially in children over the past two to three decades. What is sometimes not appreciated, however, is that increasing incidence rates of brain cancer do not necessarily equate to increasing occurrence of brain cancer.

As many fear, brain cancer trends could be reflecting increased exposure to etiologically relevant environmental toxins or hazardous life style choices. They could, however. also reflect changes in a variety of factors related to improved case ascertainment. In a recently published report [15], Malcolm Smith and several colleagues from the U.S. National Cancer Institute explored the latter conjecture. They conducted a statistical analysis of childhood brain cancer incidence rates using, national data from the population-based SEER [1] cancer surveillance system. Their a priori hypothesis was that the introduction and dissemination of magnetic resonance imaging (MRI) technology in the mid-1980s resulted in improved detection and reporting of pediatric brain cancer. thus largely explaining the observed U.S. trends. They did not average yearly rate changes over the 1973-1994 time period of their study using a linear model with a constant annual rate of increase, as others have done. Rather, they accounted for a sharp increase in rates that occurred in 1984 and 1985 using a step function ("jump") model. The jump model provided a statistically significantly better fit to the temporal incidence data than did the linear model. Their results showed that incidence rates of childhood brain cancer varied little from 1973-1984, a jump in rates occurred in 1984-1985, and a new baseline rate was established after 1985 that remained essentially stable through 1994. In other words, the average annual increase in childhood brain cancer rates that was calculated in previous reports using linear models was explained primarily by the sharp rate increases that occurred in 1984 and 1985.

Figure 1 illustrates childhood brain cancer trends based on SEER data using the modeling approach employed by Smith et al. [15]. The estimated average percentage change (EAPC) from 1975-1995 using a linear model was +1.5% per year. The jump model provided a better fit than the linear model (P = 0.003) and showed that the EAPCs froth 1975-1984 and 1986-1995 were both -0.1 %, respectively. These data are consistent with stable age-adjusted annual rates of 24 per million children in the earlier period, and 30 per million children in the later period.

The Smith et al. report [15], and an accompanying editorial by William Black [16], presented evidence in support of the MRI-detection effect, in addition to the statistical analysis. This included: 1) the timing of MRI introduction in the U.S.; 2) the rapid nature of the diffusion and application of MRI technology in the U.S.; 3) the fact that focal low grade malignancies, especially of the brain stem and cerebrum for which MRI has considerable detection benefits over CT scans, substantially accounted for the increased rates; and 4) that mortality rates did not mirror the increase in incidence rates, despite the lack of substantial improvement in brain cancer survival over the same time period. Changes in the mid-1980s in other diagnostic capabilities, such as stereotactic biopsy, may also have accounted for some portion of the observed rise in incidence rates.

Fig. 1. -- Average annual incidence rates of brain cancer in children younger than 15 years in the United States. SEER 1975-1995. Rates were adjusted to the 1970 U.S. standard population. From 1975-1995, the rates increased an average of 1.5% per year (long continuous line). The average rate change for both periods 1975-1984 and 1986-1995 (short lines) was -0.1% per year. The incidence rates averaged 24 per million children froth 1975-1984, and 30 per trillion children from 1986-1995.

The report by Smith et al. [15] lends strong support to the contention that recent increases of childhood brain cancer incidence rates in the U.S. are due, at least to a large extent, to improved detection and reporting coincident with the advent of MRI in the mid-1980s. If correct, then some consolation can be taken that the rise in rates did not likely represent an increase iii the exposure to environmental or behavioral hazards resulting in more brain cancers in children. This consolation must be balanced, however, by our continuing lack of knowledge about what specific factors are causally related to childhood brain cancer [ 17, l8]. This remains the case despite recent research efforts to reveal the role of likely candidates [19-25].

The controversy related to the increasing childhood brain cancer rates points out an important issue: Interpretation of temporal trends is often a tricky business. An analysis of incidence rates over time will reflect not only true increases or decreases. Also to be considered are temporal changes in population characteristics, the accuracy of census estimates, screening practices, diagnostic technology, morphology classifications, and other practices influencing case ascertainment. One or more of these elements can effectively conspire to show increasing incidence rates over time that is not reflective oh more cancer, but rather of better case identification. Because of the relative rarity of childhood cancers, especially when stratified by site or histology, incidence rates are considerably more sensitive to changes in case identification than are most adult cancers.

The difficulty in properly understanding the contribution of confounding factors when looking for true changes in disease occurrence. however, certainly does not mean surveillance efforts should be stopped. Public health surveillance systems are essential tools for quantifying and tracking changes in the burden of disease within a defined population. Rather, the example cited here serves to reinforce the need for a cautions and thoughtful approach to the interpretation of all data. As Black points out in his editorial [16], the report by Smith et al. [15] should lead us to other questions that need to be answered in relation to childhood brain cancer. Black includes these: "How many children have unrecognized disease, what is the full spectrum of the disease, how should the disease be managed. and what signs or symptoms should prompt an MRI examination in the first place?" I would add that we also need concerted research efforts toward identifying both exogenous and genetic factors that increase the susceptibility to brain cancer in children.

Parts of this article were published previously in the Children's Cancer Research Fund's "C3 Causes of Childhood Cancer Newsletter" (Vol. 9, No. 5, October 1998; Oie@epivax.epi.umn.edu  The author thanks Lynn Ries and Malcolm Smith of the U.S. National Cancer Institute for providing the information contained in Figure 1.

1. Ries LAG. Kosary CL, Hankey BF, et al., editors. SEER cancer statistics review, 1973-1995. Bethesda, MD: National Cancer Institute; 1998.

2. Gurney JG, Wall DA, Jukich PJ, et al. The contribution of nonmalignant tumors to CNS tumor incidence rates among children in tile United States. Cancer Causes Control (in press).

3. Cushman J. U.S. reshaping cancer strategy as incidence in children rises: increase may be tied to new chemicals in the environment. The New York Times, September 29, 1997, p A1.

4. Burcum J. A medical enigma: a rise in brain tumors sets off search for a reason. The Minneapolis Star Tribune, January 6. 1999, p A12.

5. Larsen NS. Brain tumor incidence rising: researchers ask why (news). J Natl Cancer Inst 1993:85:1024-1025.

6. Davis DL. Ahlbom A. Hoel D, et al. Is brain cancer mortality increasing in industrial countries'? Am J Ind Med 1991;19:421-4:1.

7. Workshop Group on Brain Cancer. Increasing incidence of brain cancer. Can Med Assoc J 1992;146:847-848.

8. U.S. EPA Conference on Preventable Causes of Cancer in Children (minutes). Washington, DC: U.S. Environmental Protection Agency. Office of Children's Health Protection; September 1516. 1997.

9. Gurney JG. Davis S, Severson RK. et al. Trends in cancer incidence in children in the U.S. Cancer 1996;78:532-541.

10. Bunin GR. Feuer EJ, Witman PA, et al. Increasing incidence of childhood cancer: report of 20 years experience from the Greater Delaware Valley Pediatric Tumor Registry. Paediatr Perinatal Epidemiol 1996:10:319-338.

11. Mail V. Birch .1M. Patterns and temporal trends in the incidence of malignant disease in children: II. Solid tumours of childhood. Eur J Cancer 1994:30A:1498-151 1.

12 McKinney PA, Ironside JW, Harkness EF, et al. Registration quality and descriptive epidemiology of childhood brain tumours in Scotland, 1975-90. Br J Cancer 199.1:70:07.3 979.

13 Kaatsch P, Haaf G. Michaelis J. Childhood malignancies in Germany-methods and results of a nationwide registry. Eur J Cancer 1995;31 A:993-999.

14. McWhirter WR, Petroeschevsky AI.. Incidence trends in childhood cancer in Queensland, 1973-1988. Med J Aust 1991;154: 453-455.

15. Smith MA, Feidlin B, Ries LAG, et al. Trends in reported incidence of primary malignant brain tumors in children in the United States. J Natl Cancer Inst 1998;90:1269-1277.

16. Black WC. Increasing incidence of childhood primary malignant brain tumors -- enigma or no-brainer' J Natl Cancer Inst 1998;90: 1249-1251.

17. Davis FG, Preston-Martin S. Epidemiology, incidence and survival. In: Binger DD, McLendon RE. Bruner .IM. editors. Russell & Rubinstein's pathology of tumours of the nervous system. Vol I, 6th ed. London: Arnold, 1998, p 5-145.

18. Kuijten RR, Bunin GR. Risk factors for childhood brain tumors. Cancer Epidemiol Biomarkers Prev 1993:2:277-288.

19. Preston-Martin S, Pogoda JM, Mueller BA, et al. Maternal consumption of cured meats and vitamins in relation to pediatric is brain tumors. Cancer Epidemiol Biomarkers Prev 1996;5:599 605.

20. Gurney JG, Mueller BA, Davis S, et al. Childhood brain tumor occurrence in relation to residential power line configurations. electric heating sources, and electric appliance use. Am J Epidemiol 1996;143:120-128.

21. Bunin GR, Kuijten RR. Buckley JD. et al. Relation between maternal diet and subsequent primitive neurectodermal brain tumors in young children. N Engl J Med 1993;329:536-5-41.

22. Bunin GR, Buckley JD, Boesel CP, et al. Risk (actors for astrocytic glioma and primitive neuroectodermal tumor of the brain in young children: a report from the Children's Cancer Group. Cancer Epidemiol Biomarkers Prev 199.1;3:197--204.

23. Pogoda JM, Preston-Martin S. Household pesticides and risk of pediatric brain tumors. Environ Health Perspect 1997:1(15:121-41229.

24. Gurney JG, Mueller BA, Preston-Martin S. et al. A study of pediatric brain tumors and their association with epilepsy and anticonvulsant use. Neuroepidemiology 1997;16:248-256.

25. Biegel JA. Genetics of pediatric central nervous system tumors. J Pediat Hematol Oncol 1997;19:492-501.