REVIEW Hiroko Ohgaki �� Paul Kleihues Epidemiology and etiology of gliomas Received: 27 October 2004/ Accepted: 1 November 2004/Published online: 1 February 2005 �� Springer-Verlag 2005 Abstract Gliomas of astrocytic, oligodendroglial and ependymal origin account for more than 70% of all brain tumors. The most frequent (65%) and most malignant histological type is the glioblastoma. Since the introduction of computerized tomography and magnetic resonance imaging, the incidence rates of brain tumors have been rather stable, with a tendency of higher rates in highly developed, industrialized countries. Some re- ports indicate that Caucasians have higher incidence than black or Asian populations, but to some extent, this may reflect socio-economic differences and under- ascertainment in some regions, rather than a significant difference in genetic susceptibility. With the exception of pilocytic astrocytomas, the prognosis of glioma patients is still poor. Less than 3% of glioblastoma patients are still alive at 5 years after diagnosis, higher age being the most significant predictor of poor outcome. Brain tu- mors are a component of several inherited tumor syn- dromes, but the prevalence of these syndromes is very low. Several occupations, environmental carcinogens, and diet (N-nitroso compounds) have been reported to be associated with an elevated glioma risk, but the only environmental factor unequivocally associated with an increased risk of brain tumors, including gliomas, is therapeutic X-irradiation. In particular, children treated with X-irradiation for acute lymphoblastic leukemia show a significantly elevated risk of developing gliomas and primitive neuroectodermal tumor (PNET), often within 10 years after therapy. TP53 mutations are fre- quent in low-grade gliomas and secondary glioblastomas derived therefrom. Approximately 60% of mutations are located in the hot spot codons 248 and 273, and the majority of these are G:C fi A:T transitions at CpG sites. TP53 mutations are significantly more frequent in low-grade astrocytomas with promoter methylation of the O6-methylguanine-DNA methyltransferase repair gene, suggesting that, in addition to deamination of 5-methylcytosine, exogenous or endogenous alkylation in the O6 position of guanine may contribute to the formation of these mutations. Keywords Glioma �� Incidence �� Survival �� Occupation �� N-Nitroso compounds Introduction Epidemiological studies form the basis of our under- standing of the human cancer burden, and are thus an indispensable tool for public health interventions. This is particularly true for regional time trends in incidence, mortality and prevalence. In addition, analytical epide- miology is fundamental in the identification of cancer risks and the assessment of their contribution to the development of human neoplasms. Genetic epidemiol- ogy has recently become an important tool for linking human cancers with specific exposures. In the past, promutagenic DNA adducts were thought to be caused only by exposure to environmental carcinogens, but it is now clear that somatic mutations may result from endogenous modification of DNA bases, particularly through reactions mediated by oxidative stress. Brain tumor incidence Descriptive epidemiology largely depends on popula- tion-based cancer registries, which record cases according to the International Classification of Dis- eases for Oncology (ICD-O). The recently published third edition [41] contains codes that closely correspond H. Ohgaki (&) International Agency for Research on Cancer, 150 cours Albert Thomas, 69372 Lyon, France E-mail: ohgaki@iarc.fr Tel.: +33-472-738534 Fax: +33-472-738564 P. Kleihues Department of Pathology, University of Zurich, Zurich, Switzerland Acta Neuropathol (2005) 109: 93���108 DOI 10.1007/s00401-005-0991-y

to the current histopathological classification [68], and is expected to greatly facilitate epidemiological analyses of nervous system neoplasms. In addition to tumors of the brain (ICD-O location codes 191.0���191.9), spinal cord (192.2) and cranial nerves (192.0), registries may add tumors of the meninges (192.1 and 192.3). Often, incidence data include only malignant neoplasms (last digit of the ICD-O code /3) and thus exclude most meningiomas and, previously, also low-grade astrocy- tomas. Since the latter eventually have a fatal outcome, all astrocytic tumors now have the ICD-O extension /3, with the exception of pilocytic astrocytoma. The Cen- tral Brain Tumor Registry of the United States (CBTRUS) has successfully promoted the inclusion of all histological subtypes, including benign lesions, in the SEER database [20, 27]. The age-adjusted incidence rates of brain tumors generally tend to be highest in developed, industrial countries [39, 106]. In western Europe, North America, and Australia, there are about 6���11 new cases of pri- mary intracranial tumors (including meningiomas) per 100,000 population per year in men and 4���11 new cases in women [39, 106] (Fig. 1). Even within Europe, inci- dence rates differ between countries (Fig. 2) [38]. The lower incidence in developing countries may be partly due to under-ascertainment [134], but ethnic differences in susceptibility to development of brain tumors cannot be excluded. Caucasians are more frequently affected than people of African or Asian descent [72, 87], and this difference has also been observed in children [134]. An ethnic analysis of 8947 cases of primary central nervous system (CNS) tumors registered during 1971���1985 at the Armed Forces Institute of Pathology (AFIP), Washington, DC showed that gliomas were twice as frequent in whites than in blacks [37]. Glioblastomas and germ-cell tumors were 3.5 times more frequent in white than in African Ameri- cans [28]. Similarly, CBTRUS data indicate that the incidence rates of glioblastomas, other gliomas, and germ-cell tumors are approximately twice as high in whites as in blacks, whereas the incidence of meningio- mas, lymphomas, and tumors of the sellar region (pitui- tary tumors and craniopharyngioma) did not differ significantly [20]. To some extent, these studies may re- flect differences in socioeconomic status rather than in genetic susceptibility [37]. A population-based study on primary intracranial tumors in Kumamoto prefecture suggests that, in Japan, gliomas are about half as frequent as in the USA [72]. Incidence rates of astrocytic tumors in Fig. 1 Global incidence and mortality rates of nervous system tumors, adjusted to the World Standard Population (all ages per 100,000 persons per year) [39]. Rates tend to be higher in highly developed countries 94

Asian populations are generally lower than those in Europe and North America [106] (Fig. 3). Recent popu- lation-based incidence rates from Switzerland and the USA with histological stratification are shown in Table 1. There has been some controversy regarding a possible increase of approximately 1���2% per year in the inci- dence of brain tumors [90, 112] during the 1980s and 1990s, particularly in the elderly [9, 44, 129], but also in children [136]. This apparent increase appears to be Fig. 2 Incidence and mortality rates (per 100,000 persons per year) of nervous system tumors in Europe, adjusted to the European Standard Population [38] Fig. 3 Incidence rates of astrocytic brain tumors in various countries and world regions adjusted to the World Standard Population (all ages per 100,000 persons per year). Asian populations generally show lower incidence rates than Caucasians [106] 95